MACHINE LEARNING APPLICATIONS EFFICACY AND ROLE IN HEALTH CARE PREDICTION AND SAFETY A CASE OF COVID SITUATION IN UK
Table of Contents
- Introduction 4
2.1 Role of machine learning in healthcare during COVID situation 6
2.2 Digital Technologies in Healthcare During Covid-19 7
2.3 Involvement of machine learning in healthcare decision making 8
2.4 Machine learning-based prediction of COVID-19 diagnosis based on symptoms 8
2.5 Significant Applications of Machine Learning for COVID-19 Pandemic 9
2.6 Machine learning in LPWAN oriented IoT healthcare applications 10
3.1 Overview of the Methodology 14
3.9 Validity and Reliability 16
4.1 Challenges during Covid-19 Pandemic and solved by Machine learning 19
Managing Limited Healthcare Resources 19
4.3 Linear Regression, Random Forest and Decision Tree Model 24
1. Introduction
The covid-19 pandemic provides a threat to the healthcare industry of the UK. In this global urgency situation, the UK healthcare industry needs to implement innovative technologies to solve this problem. It is essential to incorporate machine learning technology to analyze vital information regarding the UK healthcare industry after collecting data from various resources. Machine learning technology encourages researchers to find a new angle for fighting against novel coronavirus outbreaks (Majhi et al., 2020). It helps in better utilization of healthcare resources and provides a better patient management plan. Healthcare organizations of the UK can implement machine learning-based systems for developing personalized patient management plans that reduce the UK’s mortality rate due to the covid-19 pandemic. Apart from this, machine learning technology is also able to conduct covid-19 prediction using an appropriate algorithm (Sadik et al., 2020). Multiple machine learning algorithms are present that help in covid-19 prediction, such as random forest algorithm, decision tree algorithm, regression algorithm, Polynomial Regression algorithm, Long Short-Term Memory algorithm, Multilayer Perceptron algorithm, etc. These algorithms play a vital role in predicting the number of positive cases with minimal error.
1.1 Research aim
The main aim of this research paper is to incorporate multiple machine learning algorithms to identify positive cases in the UK and implement this innovative technology to enhance safety.
1.2 Research objectives
In order to achieve the research aim, it is essential to address all research objectives. The research objectives are-
- To identify current problems of the UK healthcare industry which can be solved through machine learning technology.
- To implement Polynomial Regression Model and Multilayer Perceptron Model for predicting positive cases of covid-19 in the UK.
- To develop a model based on non-linear regression, random forest, and decision tree algorithm for predicting covid-19 positive cases.
- To enhance the safety system of the UK healthcare industry using machine learning technology.
1.3 Research question
- What is the current challenge within the UK healthcare industry during the Covid crisis that can be solved using machine learning applications?
- What are the roles of the Polynomial Regression Model and Multilayer Perceptron Model to predict the COVID-19 positive cases in the UK?
- How can linear regression, random forest, and decision tree algorithms make a model that would predict positive cases due to COVID-19?
- What is the application of machine learning predictive models detecting patterns oriented with health conditions and chronic disease?
1.4 Research scope
This research regarding the application of machine learning’s efficacy and its role within the COVID-19 disease prediction in healthcare of the United Kingdom has a vast scope to be further implemented. If an individual is diagnosed and confirmed as a case of COVID-19, then the subsequent essential phase would be tracing the prevention to solve the widespread disease. Here machine learning algorithm-oriented applications or their associated approaches could be quite beneficial for the healthcare sectors of the United Kingdom so that the COVID-19 disease might be controlled. The active process of contact tracing is generally to signify and manage the individual who has been exposed to the infectious disease due to COVID-19 to reduce the spread further quite appropriately. These machine learning algorithms and their practical approaches have been enormously becoming an essential part of the advanced healthcare of the United Kingdom. These machine learning applications have been applied in different areas of healthcare during the COVID-19 crisis, such as semantic segmentation, super-resolution, multimedia system, image classification, disease recognition, and so on.
1.5 Research rationale
The core rationale behind this research is, showing the way machine learning algorithms, approaches, or applications can be proven as quite an essential factor to be chosen to fight the COVID-19 disease at the UK-based healthcare. These machine learning applications will enable the smart healthcare services that will need different processes such as preventing, measuring, and managing the COVID-19 widespread.
2. Literature review
The rising availability of electronic health data and information has presented an enormous opportunity to healthcare for both practical applications and their discovery to develop healthcare further. Here the fact is, for the healthcare sectors’ epidemiologists during the COVID-19 crisis to best utilize the data and information, the computational technique is considered essential and will be able to handle the critical datasets that are necessary. Here, the ML or Machine Learning algorithm can be mentioned as the best method or tool to identify possible patterns within the data and information that can quickly provide necessary help or support. The appropriate ML applications, algorithms, or approaches will work to further transform the satisfaction of patient risk to the data and information premises quite widely in both the fields of infectious diseases like COVID and in the field of medicine. In turn, this will further decrease the spread of pathogens based on healthcare.
2.1 Role of machine learning in healthcare during COVID situation
According to (Alimadadi et al., 2020), the Machine Learning algorithm will work as a method or tool in order to identify the patterns of data and information regarding patients’ data and other healthcare data. This particular machine learning algorithm can also be utilized in order to either boost or comprehend the recent COVID-19 disease worldwide through Identification of the possible infection risk factors as well as making predictions for the future regarding the people who will get affected (Alimadadi et al., 2020). The Machine Learning algorithm or approach will draw different factors to the healthcare field, such as statistics, science, and optimization. Apart from that, each and every Machine Learning problem will also be formulated as the only optimization challenge in respect to the core dataset. In this setting, the objectives are to find the best Machine Learning model that will explain the healthcare data at the time of the COVID- 19 situations. There are diverse types of Machine Learning algorithms such as unsupervised, supervised, and reinforcement learning.
However, (Bachtiger 2020) argued that this is partially true as in supervised learning, the Machine Learning setting will merely work on the interest-based on whether the patients are not infected or infected due to the COVID-19 virus (Bachtiger, 2020). Here the Machine Learning algorithm will map the different sets of covariates, including the patients’ demographics, to the result. This phase will further be effectively performed upon the training data and information. After learning, the Machine Learning mapping will further be applied to the core updated test data. Furthermore (Walsh, 2020) added that this is for either recognition or predicting possible deeds. As an epitome, given the primary datasets of the different patients elaborated by both the admission details and demographics, the users will endeavor to predict the main result of the readmission of thirty days (Walsh, 2020). However, the fact is, numerous learning algorithms of Machine Learning generally exist in order to accomplish further the deeds such as decision trees, logistic regression, deep neural networks, and ensemble approaches. But (Mashwama, 2021) perceives that machine learning does not contain a general algorithm as it has also experienced a large number of developments in the latest years as well as having been continuing to provide an impact over multiple disciplines quite appropriately. For multiple images identifying deeds, the core performance of the Machine Learning algorithm has been proven as entirely beneficial to healthcare during the COVID-19 crisis.
2.2 Digital Technologies in Healthcare During Covid-19
As stated by (Budd et al.,2020), digital technology has been improving tenfold during the last decade. Specifically speaking, the advent of machine learning and its algorithms have improved a lot in the last ten years. The improvement of machine learning is also helping in healthcare, and in the current situation of the pandemic covid-19, which has spread worldwide, the machine learning requirements in the healthcare industry have helped from manufacturing to physician help. Machine Learning also has increased the compilation and analysis of the COVID-19 data related to the patient worldwide. According to the author (Budd et al.,2020), the advancement in the field of machine learning has increased the efficiency of many industries, and healthcare is one of them. Here (Bayat, 2020) argued that some healthcare sectors in the UK still did not implement machine learning that includes surveillance of the population, Identification of COVID-19 cases, tracing the contact of the victims, and evaluation of interventions of public health. Machine learning also helps to compile and analyse large data sets of different healthcare centers to provide an accurate rate of COVID-19 spread in a country (Bayat, 2020). Besides this, the author (Abir., 2020) has proposed a specific method and argued against this method.
According to (Abir et al., 2020), different applications of Artificial Intelligence play an important role in detecting infectious disease and brings a revolution in the medical sector. On the other hand, machine learning has the ability to help the work of every medical professional by examining a huge amount of patient data from the medical records.
2.3 Involvement of machine learning in healthcare decision making
According to the author (Jayatilake and Ganegoda, 2021), nowadays, diseases must be identified early or detected as early as possible, such as Covid-19. These necessities lead to the incorporation of machine learning in health care provision. Machine learning helps to analyse complex medical data, medical images, and medical reports in much less time and with great accuracy. Here (Lee, 2020) also added that, as the worry grows, it is being redirected toward treating patients not only based on the sort of disease they have but also on their genetics, a process known as precision medicine. Machine learning algorithms are being modified and tested daily to increase their performance in analysing and delivering more accurate data. Machine learning has been used in healthcare, from extracting medical documents to predicting or diagnosing diseases (Lee, 2020). The incorporation of machine learning methods into computational biology has substantially enhanced medical imaging. Many disease diagnoses are being carried out using medical image processing. Additionally (Yeşilkanat, 2020) argued that machine learning-based computational decision-making is used in patient care, resource allocation, and research on treatments for various diseases and other sectors. With the knowledge gained, it was clear that neural network-based deep learning approaches worked very well in computational biology, owing to the high processing capacity of modern and sophisticated computers, and are widely used due to their high prediction accuracy and reliability (Yeşilkanat,2020). When considering the larger picture and combining the data, it becomes clear that computational biology and biomedicine-based decision-making in healthcare have become more reliant on machine learning algorithms and hence cannot be isolated from the area of artificial intelligence.
2.4 Machine learning-based prediction of COVID-19 diagnosis based on symptoms
As presented by (Zoabi et al. 2021), efficient COVID-19 screening allows for a speedy and accurate diagnosis of COVID-19, reducing the load on healthcare systems. According to the author (Zoabi et al., 2021), prediction models that integrate many variables have been developed to evaluate the risk of infection. These are intended to aid medical personnel worldwide in triaging patients, particularly in areas where healthcare resources are scarce. These are intended to aid medical personnel worldwide in triaging patients, particularly in areas where healthcare resources are scarce. It has developed a machine-learning algorithm that was trained on the records of 51,831 people who had been tested (4769 are diagnosed positive). The test set included data from 47,401 people who were tested the next week, with 3624 of them being verified to have COVID-19. With only eight binary parameters, our model predicted COVID-19 test results with great accuracy: age 60 years, sex, known contact with an infected individual, and the presence of five initial clinical signs. Overall, we created a model that detects COVID-19 cases using simple variables available by asking basic questions based on nationwide data publicly released by the Israeli Ministry of Health. When testing resources are limited, our approach can be used to prioritize testing for COVID-19, among other things.
But (Abdallah, 2019) argued that there is a high development cost for machine learning approaches that some UK-based healthcare organizations cannot afford. Specific ethical implications are to be maintained by the healthcare industry of the UK.
However, various models were built to predict the covid-19 data by utilizing Gradient-boosting machines that are created with decision-tree base-learners. The technology gradient boosting is one of the most advanced techniques to predict tabular data, and most of the successful machine learning algorithms use the Gradient boosting concept. The missing values were naturally handled by the predictor of the Gradient boosting techniques. The model uses a Gradient boosting predictor model that is trained with LightGBM with packages of python. The set of validation is used for the early alert of prevention with auROC and used for the performance measures. The Identification of the features driving model of the prediction, SHAP values that were calculated. The values are suited for complex modelings like Gradient boosting and artificial neural networks. SHAPE has the origin of the game theory, which values the partition of the prediction driving model that was calculated for the Identification. The values that were suited for the complex model of the artificial intelligence models. This is accomplished by the estimation differences between models with subsets of the feature space. The SHAP calculates its values by estimation of the contribution of each individual feature for the overall prediction.
2.5 Significant Applications of Machine Learning for COVID-19 Pandemic
According to (Mahmood, 2020), Machine learning plays an essential role in identifying all the collected cases. It also helps to detect the location where disease can be spread in the future. Machine learning technology has the ability to determine different characteristics of the disease through the collection of numerous data from social media. Moreover, Machine learning technology has the ability to correctly predict different infected cases and deaths in any particular area.
On the other hand, (Magdon-Ismail, 2020) has argued against the previous concept of the machine learning application in the COVID-19 pandemic situation.
As per the view of (Magdon-Ismail, 2020), the main objective of using machine learning in order to collect different actionable insights for developing the understanding of community health, these understandings are the probability of the transmission as well as the optimum rate at which the mild infection can be transformed into a chronic infection.
Furthermore, (Ghoshal, 2020) has argued with this idea and proposed a specific condition in order to detect the disease. According to Ghoshal et al., Bayesian Convolutional Neural Networks depend on the Monte-Carlo dropout to examine the uncertainty in different deep learning methods. These deep learning methods are utilized in order to develop the performance of the detection process at the time of checking the X-ray reports of the chest of COVID-19 affected patients. This author has concluded that if this proposed model is responsive to the uncertainty of the systems, then this proposed method can be able to enhance the performance massively. Moreover, this method has the ability to properly facilitate the medical sector along with the broader implementation of different advanced technologies like Artificial Intelligence. Besides this (Ndiaye, 2020) has provided a different approach and argued against this concept.
According to (Ndiaye, 2020), a specific machine learning method helped the SIR model to investigate the COVID-19 pandemic properly. This author has used the traditional Kermack-Mckendrick SIR prototype in order to define the patterns of spreading of the COVID-19 pandemic. This SIR prototype has the ability to provide the way of infection transmission across the population. Besides this, the author (Hassania., 2021) has proposed different aspects against this method.
According to (Hassania., 2021), Different features of the deep convolutional network have been utilized in order to extract the most representative features utilizing state-of-the-art CNN descriptors. The employed approach has the ability to properly differentiate between healthy subjects and COVID-19 from the images of CT scan and chest X-ray in order to produce higher accuracy.
2.6 Machine learning in LPWAN oriented IoT healthcare applications
Again (Teredesai, 2018) mentioned that there are multiple healthcare applications that can further utilize machine learning and LPWAN quite appropriately. There are various devices that have been leading to an innovative improvement: the formulation of the LPWAN Oriented in the Machine learning models. Machine learning techniques are being utilized in healthcare for computational decision-making in cases where a critical data analysis on medical data is required to identify hidden linkages or anomalies that are not evident to humans. Implementing algorithms to execute such jobs is challenging in and of itself, but increasing the accuracy makes it considerably more difficult. But this is partially true as (Albeshri, 2020) added, processing vast amounts of medical data was a significant task in the early days, which led to the adaptation of machine learning in the biological domain. Since then, the biology and biomedical areas have progressed to new heights by discovering new knowledge and identifying previously unknown linkages. There are certain areas where this LPWAN oriented IoT applications considering the machine learning has been utilizing during the Covid 19 period as stated below such as:
Monitoring Temperature
According to (Khan, 2021), the appropriate usage of this LPWAN oriented IoT application considering machine learning has been supporting the healthcare professionals of the United Kingdom in order to monitor the temperature of the patients and record the data (Khan et al., 2021). One of the most critical signs of this temperature of the critically ill patients’ body is that it requires everyday monitoring. But (Abbasi, 2021) has shown a different side that the TelosB has been utilizing with the temperature sensor where the body temperature monitoring has been done merely periodically (Khan et al, 2021). As an epitome, this technology detects the peoples’ body temperature as well as transmits utilizing the infrared sensors through which this can be known that whether a person is Covid affected or not. There is also another system that has also been utilized is the RFID module. This module has also been used for body temperature monitoring by UK-based healthcare professionals.
Monitoring Blood Oxygen Saturation Level
As stated by (Kim, 2021), the monitoring of Oxygen saturation levels has also been carried out by the pulse oximeter. Here in order to perceive the core healthcare IoT applications, this pulse oximeter has also significantly been integrated with the LPWAN quite appropriately. There is a potential in order to utilize the technique that can largely be observed in the UK-based healthcare sectors. But (Khan, 2021) has presented another perspective that there are certain remote monitoring systems considering the low power of pulse oximeters situated within the IoT network of the healthcare sectors of the United Kingdom (Khan et al., 2021). Apart from that, telemedicine applications have also been using this pulse oximeter where each and every data obtained from this application have been appropriately utilized by the core machine learning software technology (Khan et al., 2021). This is significant to know regarding the possible health-oriented challenges within the monitored patients.
Managing medication
As described by (Abbasi, 2021), the LPWAN oriented IoT applications considering the machine learning has also been providing the innovative mechanism in order to manage the problem arising from non-compliance of the medication management that have been resulting critical challenge to the UK based healthcare sectors at the time of the Covid 19 crisis (Khan et al., 2021). As an epitome, this mechanism has been successfully providing the IoT-oriented smart packaging solution considering the controlled sealing based on wireless communication. Here (Kim, 2021) argued that there is also another system of the medication that has also been largely utilized that is known as the RFID, as it utilizes the prototype implementation for the system of medication control quite appropriately (Khan et al., 2021). Here the prior data and information are significantly utilizing the IoT network within the UK-based healthcare sectors in order to have better control over the core future movement of the patients’ medication utilizing the application of machine learning.
Managing rehabilitation
As discussed by (Khan, 2021), within this Covid 19 crisis, there was a core need of providing rehabilitation to the patients or the common people who have been staying in the home during the lockdown. It needs the appropriate management system to control the rehabilitation system through which they can be tracked by the UK-based healthcare sectors quite appropriately. This has also been considered as the crucial branch of the UK’s medicine in order to significantly restore the main functional ability of the patients who are having physical disabilities in the country (Khan et al., 2021). This field has been conjugated with the LPWAN, and this LPWAN has been utilized appropriately in order to collect multiple data and information associated with the patients’ age, gender, the reason for taking rehab, and so on. Here (Kim, 2021) argued that this particular plan has merely been obtained with the machine learning applications for having the data and information of the earlier patients as well (Khan et al., 2021). As an epitome, these systems have been helping in order to collect the patients’ data for the purpose of rehabilitation utilizing the UK-based healthcare sectors’ IoT network where they can be utilized in order to make the appropriate rehabilitation plan with the machine learning applications.
Healthcare solution based on smartphone
As stated by (Kim, 2021), during this Covid 19 crisis, the healthcare professionals situated within the UK have become quite unable to provide treatment to each and every patient who is suffering from the Covid 19 or not. Therefore, the UK-based healthcare sectors have prepared to launch smartphone-oriented healthcare solutions for the multifarious patients in the country so that each patient can be treated either physically or virtually in the country. This is launched considering the approval of the UK Government (Khan et al., 2021). The fact is, this smartphone has become a necessary gadget, and this merely can be the only solution to mitigate the raised challenges from the Covid 19 crisis. There are myriad healthcare IoT-based applications that have been introduced to benefit both the healthcare services and the patients of the country. But (Khan, 2021) argued that there are merely some of the smartphone-oriented healthcare accessories, named Fitbit Flex, this has been utilizing for multifarious individuals so that they can be able to achieve a better fitness level to fight the Covid 19 virus and stronger the immune system (Khan et al., 2021). These smartphone-based IoT healthcare applications have been strongly integrated with the backend system in order to gather further the necessary data and information of the patients considering the machine learning algorithm.
Testing Blood Glucose
According to (Abbasi, 2021), Blood Glucose Monitoring has been considered the crucial practice for diabetic patients, and they generally have a high level of glucose over an extended period. This particular monitoring can indicate the best pattern regarding the glucose level present in the patient’s blood. This also further helps in deciding the medications, meal plans, and other physical activities that are needed for the patients (Khan et al., 2021). At the time of the Covid 19 crisis, this machine learning-oriented LPWAN oriented IoT healthcare applications have been significantly utilized. These LPWAN oriented IoT applications considering machine learning have also become beneficial in order to further decide the appropriate Covid 19 treatment plan depending upon the patients’ prior history (Khan et al, 2021). But here (Khan, 2021) argued that, testing blood glucose level by utilizing this technology will not solve the raised problem due to Covid 19 virus. Apart from that these technologies have also been collecting the sensitive and confidential data of the patients. Here the healthcare providers have been given access regarding the patients’ data and information collecting system through which they can be able to prepare for the subsequent treatment of Covid 19.
Methodology
In order to accomplish this research paper, the authors will follow a secondary research methodology. This research methodology collects secondary data from previously published research papers, journals, and articles based on machine learning technology and the UK healthcare industry in the current covid-19 pandemic situation (Alotaibi et al., 2020). The authors choose this methodology to gather relevant research papers in less time. It is practical to select previously published research papers from google scholar to maintain authenticity.
3.1 Overview of the Methodology
Research methodology represents the particular procedures or techniques that are primarily utilized to successfully detect, select, process, and analyse information on the research topic. The research methodology conveys the way of choosing the appropriate method so that both the process of data collection and data analysis can be successfully executed. This methodology will successfully cover research onion, philosophy, method, approach, and design. In addition, the data collection and data analysis procedure, along with research limitation, validation, and reliability, will also be discussed in the methodology section of this dissertation.
3.2 Research Onion
Research onion is considered one of the essential frameworks that provide the necessary assistance in developing a robust research methodology that can enable the researchers to conduct systematic research. It refers to an in-detailed description or discussion of several phases of the dissertation process properly. Along with that, this particular framework is also capable of delivering both an efficient progression and effectiveness, which can assist in designing the core methodology of the research. This framework is held responsible in terms of maintaining the flow of the relevant information related to the research topic. The research onions consist of different operational layers, and those layers are addressed in the picture above (Melnikovas., 2018).
3.3 Research Philosophy
At the outermost layer of the research onion framework, the section called research philosophy is located. In this particular research paper, the positivism research philosophy has been followed. This particular philosophy successfully evaluates various research scenarios with the required assistance of scientific research methods. This particular research activity is critically associated with the process of critically analysing the role of machine learning applications in the field of healthcare prediction regarding the covid-19 pandemic situation in the United Kingdom (Ahmad et al., 2018). This particular research philosophy will offer the required assistance in evaluating the dissertation’s problem statement with the employment of the secondary data. The positivism research philosophy has helped in terms of making a rewarding and genuine conclusion.
3.4 Research Method
For this particular dissertation, secondary research methodology has been chosen in order to gather multifarious information and data successfully. The secondary research methodology effectively helps the researchers in order to collect relevant information from different articles, journals, government websites, and previously published research papers that are related to this research topic. With the help of the secondary research methodology, information regarding the research topic can be collected so that the information can be properly analysed to provide relevant information to this dissertation. It can also be informed that the major eason for choosing this specific research methodology is its cost-effective and efficient nature.
3.5 Research Approach
For this particular dissertation, the deductive research approach has been followed after considering the current research activity so a proper conclusion can be delivered at the end of this dissertation in order to critically analyses the role of machine learning applications in the process of improving healthcare prediction (Ahmad et al., 2018). Along with that, the effective utilization of the applications which are based on the machine learning technology in the healthcare industries in the United Kingdom during the time of pandemic situation can also be analysed with the help of the deductive research approach (Alhashmi et al., 2019).
3.6 Research Design
Descriptive research findings have been followed for this particular research activity that can successfully provide the necessary help in developing the capacity to evaluate the problem statement. Different types of practical applications which are primarily based on machine learning technology that can enhance the process of health prediction are required to be mentioned in this particular dissertation. The descriptive research design approach is primarily based on different strategic elements which tend to possess the capacity which is required to evaluate the problem statement of the dissertation. So, it can be successfully ensured that the selection of the deductive research approach is adequately justified (Alhashmi et al., 2019).
3.7 Data Collection
For this particular dissertation, a secondary data collection method has been chosen to collect relevant information from journals, articles, and previously published research papers related to the research topic. This secondary data collection has been choses due to gathering most relevant data in a short time.
3.8 Data Analysis
Within this dissertation, qualitative data analysis has been utilized. In the literature review part, previously published relevant research papers which are highly related to the research topic have been utilized in this particular dissertation. Along with that, other research materials have also been utilized for this dissertation, such as journals, articles, governmental websites, and so on.
3.9 Validity and Reliability
Every single approach of collecting information regarding the research topic has been validated in a proper manner. For this particular dissertation, there was no sign of data modification or data tampering. The data which has been collected is only utilized for this particular dissertation. The reliability and authenticity of this dissertation have also been maintained during the research work.
3.10 Ethical Consideration
Several problems were faced while collecting relevant information for this dissertation, such as loss of data and irrelevant or incomplete data. Privacy, as well as proper confidentiality of the data, have been maintained properly under different circumstances and guidelines at the time of conducting this dissertation.
3.11 Research Limitation
The research limitations that have been faced by the researchers while gathering data for this dissertation are weak technical implementation, lack of relevant information, lack of support and funds.
3.12 Research plan
![](https://mygradepaper.com/wp-content/uploads/2023/12/C1234.jpg)
4. Discussion
4.0 Introduction
The report discusses some of the significant insights on implementing machine learning algorithms in identifying positive cases of COVID-19 in the United Kingdom. The use of digital technologies has provided tons of benefits to the healthcare sector resulting in better diagnosis and treatment. The medical sector is considered to be the one that invests in research and development for providing better treatment and services and the best care to the patients. However, in cases like a pandemic, it gets challenging to find an effective treatment. This is because it is hard to detect patients with the new virus. Thus, artificial intelligence in identifying patients infected with the virus has been seen to provide promising results. This section below describes different types of important points found from previously published research papers. The author below provides the exact structure of UK healthcare and shares two types of challenges that solve with the ML algorithms. The dissertation focuses on the UK healthcare industry, but the issues are common for all healthcare industries.
4.1 Challenges during Covid-19 Pandemic and solved by Machine learning
According to the researcher, COVID-19 is a global pandemic; it is a direct threat to the healthcare industry as well as breaks the infrastructure of the economy around the world. The pandemic increases the burden in the healthcare industry, which cannot be effectively dealt with by the medical teams. This research paper finds out the challenges that can easily be solved with the machine learning algorithm. Both the UK as well as the international community face an unbelievable amount of pressure for their limited infrastructure. Through the Machine learning algorithm, the researcher finds out different types of solutions that can change the way and satisfy the patients (Shahid et al., 2020).
Managing Limited Healthcare Resources
The UK healthcare industry or other countries’ health industries face severe scarcity of resources during pandemics like test kits, ICU beds, hospitals beds, and so on. Presently in the UK, the capacity of the hospital’s bed and ICU beds are occupied. This resource scarcity mainly occurs because the symptoms of the COVID-19 change in different waves. Some people are asymptotic, others like symptoms of flu, as well as some experience complications like fatal-multi organ failure, pneumonia, etc. (Shahid et al., 2020). As per the analysis of research papers, resources are limited, and COVID-19 risk increases day by day. The scarcity of resources in healthcare increases the problems, and the limited resources cannot deal with the Covid-19 patients. For this reason, there is a need to manage the limited resources with machine learning. Mature Machine learning algorithms solve the critical issues in the UK healthcare industry. With the help of the Electronic Health Record system, the healthcare industry easily holds the patient’s data. The machine learning algorithm solves the various data sources and predicts the accurate result, which helps mitigate the risk of developing during the Covid-19 pandemic.
Developing treatment plans and patient management, therapeutics, and vaccination for pandemics take some time to increase as well as introduce. Numbers of antivirus medicine were indeed tried during this pandemic, but for the huge number of patients, one nurse or doctor cannot handle it. The UK healthcare sector introduced a technology based on machine learning technology to solve the issues of limited nurses. The computer assistant documentation system provides the easiest way during this pandemic. The technology shares real-time clinical data and ensures consistent recommendations. Voice-to-text transmission is another feature of machine learning algorithms that can also solve patient management problems.
From this section, it is understood that machine learning is a resource that can solve the UK healthcare industry issues.
4.2 Roles of the Polynomial Regression Model and Multilayer Perceptron Model in Prediction of Covid-19 Cases
Polynomial regression refers to the form of linear regression analysis where the relationship between the dependable variable ‘y’ and the independent variable ‘x’ is modelled as the nth degree polynomial in the independent variable. This type of regression tends to act as a non-linear relationship between the corresponding conditional mean value which is represented as ‘y’, as well as the value of ‘x’. Polynomial regression is able to successfully mitigate the costs which are retired by the cost function (Ekum&Ogunsanya, 2020).
On the other hand, a multilayer perceptron (MLP) is the class of feedforward ANN (Artificial Neural Network). The multilayer perceptron successfully uses a supervised learning technique which is called backpropagation for training. Along with that, it can also distinguish data that is not separable linearly (Ekum&Ogunsanya, 2020).
The Polynomial Regression, as well as Multilayer Perceptron, can provide the necessary assistance in terms of predicting the number of infections, recoveries, along with deaths during the Covid-19 outbreak.
The applications are now emerging in a proper discipline such as viral epidemiology with the advancement of the technology which is integrated with the Machine Learning algorithms. Machine Learning is considered to be the technology that can successfully identify early contamination of viruses along with the process of development of vaccines and drugs as well as the risk assessments, detecting the protein structures of the detected viruses, and so on.
The PR (Polynomial Regression) algorithm is utilized when the independent data’s range fluctuates a lot (Ekum & Ogunsanya, 2020). Along with that, the Linear Regression algorithm is unable to predict the data pattern. This independent variable is primarily modelled as an nth dependent variable’s degree polynomial. This algorithm is preliminarily utilized in order to identify how diseases spread as well as epidemics or pandemics. The algorithm of Polynomial Regression (PR) is based on the equation which has been provided in the following.
In the equation, the ‘x’ represents the input terms, as well as the dependable variable ‘y’ is considered to be the predicted input.
The Covid-19 is now addressed as the newly emerging virus. Several datasets are accepted internationally from the IEDCR and World meter, the statistical data related to the global condition of Covid-19. This dataset contains time-series data of the number of the total recovery cases, totally confirmed cases, total death cases. These data are then integrated into the ML (Machine Learning) Models to predict the outbreak of the covid-19 pandemic (Ekum & Ogunsanya, 2020). By applying the Machine Learning algorithms, the trend of the outbreak can be predicted in an accurate way by using the memory algorithm by Multilayer Perceptron, Polynomial Regression, as well as Long Short Term. In order to successfully predict the entire confirmed deaths, cases as well as recoveries LSTM (Long Short-Term Memory), as well as PR (Polynomial Regression) prediction curve, is considered smoother than Multilayer Perceptron that implies that both Long Short-Term Memory, as well as Polynomial Regression, are well suited or appropriate for the outbreak than the Multilayer perceptron. In the next section, the role of multilayer perceptron is critically analysed in terms of predicting the Covid-19 outbreak (Ekum & Ogunsanya, 2020).
Multilayer Perceptron (MLP)
In the first section, the author discusses the polynomial regression model and its benefit to solve pandemic-related issues. Another model can effectively solve the positive case of the COVID-19 pandemic and predict the accurate result. According to the researcher, the Multilayer Perceptron model is known as an artificial intelligence neural network method that is effectively used to collect time-series data (Monrat et al., 2020). So, it is a good choice for the researcher to contain the temporal ordering data. From the analysis of the research paper, it is identified that MLP includes a hidden layer, one input layer, and one output layer. Each layer except the output layer contains a bias neuron that is fully connected to the next layer. The layering theme of the MLP is described in the figure below (Monrat et al., 2020).
The backpropagation algorithm is applicable to order the MLP network; for every training instance, the MLP algorithm computes the output of each neuron in a different layer. After that, it calculates the network output error then goes to each layer in reverse order to find out the error contribution. The computation of every layer follows a different equation.
From the above picture, it is identified that ‘W’ represents a weight between hidden layers as well as input neurons, then ‘b’ represents a bias weight. Here the author presents the output of the hidden layers (Monrat et al., 2020).
After that, the total input to the output stage at the “k th” neuron of the hidden bias is
The ‘k th’ neuron of the production value is
After applying the MLP algorithm, the researcher of the research paper finds out an output, the spread of the virus is very unpredictable, and the Covid-19 spread fluctuates along with government policies and awareness. Through the polynomial regression and MLP algorithm, it is easy to store the record of the patients and constantly update the government to control the outbreak (Xue et al., 2020).The researcher develops an SIR model based on the above algorithm’s equation and chooses the timeline of the 60 days. The below picture shows the Covid-19 outbreak. The SIR model makes it easy to predict the Covid-19 pandemic result and update the government simultaneously. The machine learning model easily trains the algorithm to grab the effective result based on the 60 days. For this reason, it is true that with the help of these two algorithms, it is easy to collect effective information during the Covid-19 pandemic (Xue et al., 2020).
4.3 Linear Regression, Random Forest and Decision Tree Model
Furthermore, the decision tree is a very effective machine learning model in terms of detecting COVID-19 infection. This model is mainly utilized in order to divide the learning tasks by splitting the dataset into different small groups until data types examine the data categorization. In this model, the division of attribute of numerical data type: {B}<–z, where z is the value of B domain for the total attribute of the data type Partition C. In this decision tree model, every path starts from the root and represents a specific data sequence of splitting data until it reaches the Boolean outcome. In the case of COVID-19 prediction, this decision can be considered due to its massive understandability and clearness (Thakre et al., 2021).
The Random Forest Classifier is very popular as an ensemble classification technique in the data science sector and machine learning field (Thakre et al., 2021). This can be considered because this model provides “parallel assembling” that includes many different decision tree classifiers in numerous data sets in parallel. This process utilizes the average or voting in order to identify the final result. On the basis of these properties, the single decision tree is quite less effective than the random first machine learning model. On the other hand, it has the ability to reduce the issues of overfitting at the time of enhancing control.
4.4 Application of machine learning predictive models
There are a number of distinct predictive models of machine learning that are very relevant to utilize in the process of chronic disease diagnosing. In this chronic disease diagnosis. The machine learning predictive model is more effective and depends on user access and data size. Among these predictive models of machine learning, the supervised machine learning model is one of the effective machine learning models because the integration process of this model is very easy (Battineni et al., 2020). Furthermore, the implementation of this model in healthcare can provide better health services by enhancing the decision-making ability of health specialists. It has been confirmed that every machine learning algorithm is based on a different particular problem as well as review studies can help in order to properly examine the effectiveness and performance of every machine learning model.
Before implementing machine learning in the healthcare industry, the development of the medical practice was based on individual studies (Chen et al., 2017). On the other hand, this machine learning model is affecting data science because all of the medical data is gathered from several platforms as well as different people. As the cause of the rapid growth of different computational models, most of the healthcare industries in the UK as well as across the world have been quickly transformed through achieving the ability to record huge amounts of patient data. It is very difficult to examine all of the medical records with human knowledge properly (Battineni et al., 2020).
Therefore, the utilization of big data in different medical services networks can help develop patient care in chronic diseases. As per the given scenario, it has been seen that the aging issue is one of the significant issues in the COVID-19 pandemic and other chronic diseases also. Most of the researchers anticipated the safety of every aging compound along with the collection of a number of deep neural network classifiers. Moreover, it can be said that predictive models of machine learning have the ability to provide active decision support. This dissertation has provided different machine learning models that can be utilized for COVID-19 detection and another chronic disease. The regression problem always looks at continuous data as well as most of the researchers have followed this in terms of gene expression (Battineni et al., 2020).
Furthermore, unsupervised Machine Learning models deal with different data sets that contain the deep learning model in order to monitor these data properly. The neural network is utilized in order to auto-detect structured data for extracting every key feature. This discussion part of this dissertation has suggested that enhancing the power of different predictive models in the diagnosis of Chronic Disease has the ability to empower all of the medical experts. This will help to bring an important tendency in terms of proper decision-making at medical centres across the world (Battineni et al., 2020).
There are several effective ways available in order to predict the connection between two variables. Linear regression is one of the most popular machine learning models. In this case of the COVID-19 prediction, the linear regression can be described in the following equation, [Y=a+bX] (Mathur et al., 2021).
The linear regression model is only effective when there is no relationship between dependent and independent variables. In this particular equation, ‘y’ is the independent variable and this can be categorical or continuous (Mathur et al., 2021). On the other hand, the ‘X’ is the dependent variable and this is the continuous value. As per the equation the line slope is ‘b,’ and the interpretation is ‘b’ {here the intercept means the value of ‘y’ when X=0}. This particular situation has been examined along with the probability distribution in order to focus on the multivariate analysis and conditional probability distribution (Mathur et al., 2021). On the other hand, an effective scatterplot can be used in order to determine the correlation strength between two different variables.
This Linear Regression can be utilized in order to predict COVID-19 cases at the time of lockdown in the United Kingdom at the time of very few rising cases.
It has also been understood that SVM (Support Vector Machine), as well as the LR models, are successfully implemented within a wide range of studies regarding the related topics in order to make CD diagnoses. The Support Vector Machine model is considered to be one of the most popular models among the other models in terms of detecting (COPD) Chronic Obstructive Pulmonary Disease from the start. Along with that, it has also been understood that it can be successfully assisted within the relationship of the patient and the doctors (Battineni et al., 2020). NB models, along with the Bayesian networks, assist in forecasting the diagnosis of patients who are suffering from asthma (Kassania et al., 2021). These particular models successfully encompass the old patient records in order to look up the footing and clinical symptoms on the Bayesian networks in terms of presenting the relationship between individual cases (Battineni et al., 2020). This will also provide the necessary assistance in terms of diagnosing possible future symptoms. In this case, the KNN (k-nearest neighbours) is primarily associated with five different studies for diagnosis, forecasting so that the stage of CD diagnosis can be successfully followed with the necessary assistance of the secondary as well as primary data. Because of that, it has also been noticed that it is very important in terms of adopting the clustering (unsupervised) as well as neural networks (deep learning) modes within the future (Battineni et al., 2020).
Chronic Diseases (CDs) are held responsible for a higher portion of international health costs. Patients with chronic diseases require lifelong treatment. In this case, the predictive models that are integrated with the machine learning technology are developed and utilized with the processes of diagnosis along with forecasting or predicting these diseases. With the help of the models integrated with machine learning algorithms, it becomes possible to improve or enhance the quality of the medical data (Kassania et al., 2021). This will also assist in mitigating the fluctuations within the patient rates as well as save the cost of medical activities. It is believed that to successfully decrease the death rate from chronic diseases, effective treatments and early detection of diseases can be considered the only solution (Kassania et al., 2021).
Machine learning methods are effectively utilized with the interpretation (computerized) of the mnemonic capacity tests for various types of analysis of chronic diseases. The models integrated with the machine learning algorithms are highly utilized to successfully investigate the diagnosis analysis at the time of comparison with other conventional methods. It is highly expected that the highest accuracy providing models can obtain large importance within the process of medical diagnosis (Battineni et al., 2020). The models are critically designed to emphasize the responsibility of the patient care quality and reduce or mitigate the medical costs. With numerous researches and observations, it has been observed that in this particular field, Support Vector Machine with the kernel methods can be addressed as one of the most useful and effective methods that can successfully detect the death rates and recovery rates of chronic diseases (Kassania et al., 2021). In the Covid-19 pandemic situation, it has also provided the necessary assistance in terms of gathering the relevant medical data that would help to analyze the entire situation and also help to develop effective strategies that can help to develop effective mitigation strategies and mitigate the future risk factors (Battineni et al., 2020).
Conclusion
It can be concluded that the healthcare industry applies several methods based on the machine learning algorithm, and that solve pandemic-related issues easily. This discussion section elaborates on different types of challenges that are solved by ML algorithms. After that, the author applies two types of algorithms in the second section, along with the calculation that elaborates the present situation of the government and how the algorithm provides the benefit through these algorithms. It is concluded that the overall discussion elaborates on the exceptional thought of the Covid-19 pandemic in UK healthcare.
5. Findings
It has been traced that the usage of innovative technology such as machine learning has successfully secured data and information sharing at the time of the COVID- 19 pandemic, which is quite effective. The standardization method of both network protocol and latest communication have been enabled by the usage of this machine learning in the healthcare sector of the United Kingdom. The machine learning models have been detecting possible attacks on the healthcare cyber security system. In addition, the machine learning algorithm has also been forecasting disease trends as well as improving the programs of social awareness that have been outlined here in brief.
The SARS-CoV2 has been a critical-ranging crisis because of the increased rate of mortality as well as a lack of medical treatment all over the world. Despite having the best healthcare infrastructure in the country, the UK has been unable to deliver 100% services to every patient. The fact is, appropriate implementation of technology such as machine learning, big data, and other crucial mechanisms can help secure the sensitive data of the patients and help in diagnosing diseases. Therefore, the researchers and the healthcare providers of the UK have been quite focused upon the machine learning approaches and its core algorithm to boost further both the speed and the way of processing performing multifarious health-oriented deeds in the country. Furthermore, the core health clinical system oriented with the machine learning algorithm has been significantly addressing the major challenges of the Covid 19 crisis.
This is also a fact that. Within this UK-based healthcare industry, the machine learning approaches will not be able to replace human interaction. Still, significantly, this can deliver the best decision support to the clinicians for modelling the sensitive and confidential data and information with further predicting the core outcome quite appropriately. These e-health applications have been utilized for multifarious healthcare services in order further to mitigate the spreading of the Covid 19 virus. In this particular situation, the UK-based researchers have also initiated to strongly focus upon the multiple cases of the IoT-based applications utilizing the technology of LPWAN. It has been found that the core combination of both the technology such as machine learning algorithm and LPWAN has been proven successful in different deeds. They have been proven in diagnosing diseases, predicting treatment, providing data security, and so on.
As the efficacy of machine learning is applicable for the prediction of COVID-19 cases and the ongoing healthcare scenario in the UK. Machine learning is considered a relevant form of Artificial Intelligence (AI) that will be effectively used in medical imaging. In the NHS (National Health Service) of the UK< it is evident that data-driven decision-making is made possible due to the usage of machine learning systems to detect the current condition of the healthcare industry and the COVID-19 situation in the UK. Speed and accuracy can be considered the relevant efficiencies and advantages of machine learning systems for depicting the current healthcare scenario with safety and the identifiable scenario of effects of COVID-19 in the UK (Bachtiger et al. 2020).
A promising screening model is considered a significant benefit for machine learning systems and the development of efficiencies. The role of interpretation of ICT is relevant for the depiction of real-world data with high levels of confidence. New algorithms may be implemented and generated during the ongoing scenario of the COVID-19 scenario in the UK and manage the complexities of the current healthcare system in the UK (Bachtiger et al., 2020).
By the application of machine learning systems to predict the healthcare scenario in the UK, along with the consideration of the COVID-19 situation, it is evident that machine learning can provide a perfect model for identifying patients and understanding the patterns of effective infection rates (Chaurasia, & Pal, 2020). Under the efficacy of machine learning, it is clear that a machine learning system will constitute the use of an Auto-Regressive Integrated Moving Average (ARIMA) model. In this model, the performance of the machine learning system can be identified effectively for the detection of the healthcare system in the UK, safety, and determine the ongoing COVID-19 situation in the UK (Chaurasia, & Pal, 2020). The machine learning system can develop a practical training module to determine the rates of infection in Mid-March (because in Mid-March, there is a rapid rise of the COVID-19 cases in the UK as well as other parts of the world) (Chaurasia, & Pal, 2020).
With the advent of COVID-19, every sector has been suffering a lot, and the healthcare sector is in great demand. In this COVID-19 situation, machine learning tends to provide quality service by forecasting the spread of disease. At the same time, the recovery and mortality rate can also be measured by it. Various machine learning approaches are available in the market, and every approach can be applicable to the industry. It has also been found that the machine learning approach also ensures the safety and security of the patients (Chen et al., 2017).
The UK healthcare industries are facing severe challenges during the Covid-19 pandemic situation in terms of allocating sufficient resources like ICU beds, test kids, and others. At the time of the peak of the pandemic situation, the capacity of the ICU beds, as well as hospital beds, were occupied in the United Kingdom (Chen et al., 2017). So, it is well understood that an effective solution is highly required to successfully deal with the problems. In this case, the applications which are associated with the machine learning algorithms can be considered to be very useful as these algorithms can successfully obtain the medical data from current hospital records regarding the number of affected patients, some patients who died or recovered (Monrat et al., 2020). This will effectively provide the necessary assistance in identifying and mitigating the issues of allocating patients within the healthcare facilities in the United Kingdom (Chen et al., 2017).
Machine learning algorithms effectively provide effective strategies that can solve different data sources along with predicting accurate results. This will also assist in decreasing the risks during the current pandemic situation in the discussion section. The role of the Polynomial Regression model and the Multilayer Perceptron model has been critically analysed in predicting the Covid-19 cases in the United Kingdom (Monrat et al., 2020). The polynomial regression model is utilized to stop the fluctuation of the range of the medical data so that relevant and reliable data can be gathered, which will be later analyzed in terms of developing effective solutions that can improve the current critical pandemic scenario by successfully increasing the recovery rates of the Covid-19 patients (Chen et al., 2017).
6. Conclusion
From this research proposal, it can be concluded that machine learning algorithms must be applied to the UK healthcare industry to predict covid-19 positive cases. Machine Learning Algorithms can successfully identify the patterns primarily associated with health conditions and diseases by successfully studying and analysing thousands of healthcare records and other patient data. After analysing previously published research papers on the role of machine learning algorithms in healthcare facilities, it has also been understood that significant developments have occurred within the field of Machine Learning-based applications. These recent developments are capable of successfully providing necessary assistance to increase healthcare access within the developing countries and innovate cancer diagnosis as well as treatment. On the other hand, reliable health forecasts or health care predictions are considered highly important in providing better health service to patients. It can be considered crucial because it can successfully improve pretentious healthcare services and develop alerts for the management or situation of the patient overflows. In addition, healthcare predictions can also successfully mitigate the associated costs in staff and supplies. The UK possesses a highly developed healthcare infrastructure, as well as the standard of healthcare is very high. Because of that, the UK ranked one of the top 10 most efficient and reliable healthcare systems all over the world.
The UK has a government-funded comprehensive health service and developing private health care sectors that successfully offer a high standard of healthcare to their patients. Within the healthcare facilities in the UK, Support Vector machines are utilized, which are considered to be one of the most standard Machine Learning (ML) algorithms. This particular machine mainly uses a supervised learning model for regression, classification as well as Identification of outlines. The dissertation chooses a secondary research methodology to grab the effective information from previously published research papers in the above portion. Also, in the methodology section, the author elaborates how effective the secondary research methodology is to grab the essential information. Different types of procedures are effective to apply in secondary research methodology procedures. For this dissertation, this type of methodology provides many advantages, such as it takes less time, is cost-effective, and so on. After that, in the discussion section, the dissertation provides a specific area of the discussion and analyses the UK healthcare system to share the huge amount of advantages in front of the patients. Discussion breaks into four stages, and each stage answers the research questions from the previously published research papers. Machine learning technology encourages researchers to find a new angle for fighting against novel coronavirus outbreaks. It helps in better utilization of healthcare resources and provides a better patient management plan. Healthcare organizations of the UK can implement machine learning-based systems for developing personalized patient management plans that reduce the UK’s mortality rate due to the covid-19 pandemic. Polynomial Regression (PR) and Multilayer Perceptron (MLP) provide a huge number of benefits in this pandemic. With the help of this algorithm, it is easy to calculate the positive cases of the covid patients and share the result in front of the government. In the finding section, the author finds that the innovative technology is useful for the healthcare industry and effectively focuses on critical patients.
The robotic system, an electronic health record system, easily grabs the patient’s information. The algorithm uses a database that is automatically updated and stores the patient records. After that, the data is visible in front of the doctors. Also, the machine learning algorithm effectively mitigates the risk of cybersecurity-related issues and solves problems from the primary stage. Lastly, it is said that the ML is needed for the present situation, but it has huge numbers of drawbacks which is complex for the healthcare industry. Sometimes the machine learning algorithm stores the poor quality of the data, which is the biggest threat for the industry; based on the poor data, the industry cannot survive. On the other hand, the technology needs to be updated several times, which is very time-consuming. When using the machine learning algorithm in the healthcare industry, it is suggested that we consciously check all the translations, and the technical team must be knowledgeable to handle the technology. Effective information provided by this algorithm also distracts the people with fake information.
7. References
Abir, S. M., Islam, S. N., Anwar, A., Mahmood, A. N., &Oo, A. M. T. (2020). Building resilience against COVID-19 pandemic using artificial intelligence, machine learning, and IoT: A survey of recent progress. IoT, 1(2), 506-528. Retrieved from https://www.mdpi.com/2624-831X/1/2/28/pdf
Ahmad, M. A., Eckert, C., &Teredesai, A. (2018, August). Interpretable machine learning in healthcare. In Proceedings of the 2018 ACM international conference on bioinformatics, computational biology, and health informatics (pp. 559-560). Retrieved from: http://www.comp.hkbu.edu.hk/~iib/2018/Aug/article1/iib_vol19no1_article1.pdf
Alimadadi, A., Aryal, S., Manandhar, I., Munroe, P. B., Joe, B., & Cheng, X. (2020). Artificial intelligence and machine learning to fight COVID-19. Retrieved fromhttps://journals.physiology.org/doi/full/10.1152/physiolgenomics.00029.2020
Alotaibi, S., Mehmood, R., Katib, I., Rana, O., &Albeshri, A. (2020). Share: A big data analytics tool for healthcare symptoms and diseases detection using Twitter, Apache Spark, and Machine Learning. Applied Sciences, 10(4), 1398. Retrieved from: https://www.mdpi.com/2076-3417/10/4/1398/pdf
Bachtiger, P., Peters, N. S., & Walsh, S. L. (2020). Machine learning for COVID-19—asking the right questions. The Lancet Digital Health, 2(8), e391-e392. Retrieved fromhttps://www.thelancet.com/journals/landig/article/PIIS2589-7500(20)30162-X/fulltext
Bachtiger, P., Peters, N. S., & Walsh, S. L. (2020). Machine learning for COVID-19—asking the right questions. The Lancet Digital Health, 2(8), e391-e392. Retrieved fromhttps://www.thelancet.com/journals/landig/article/PIIS2589-7500(20)30162-X/fulltext
Battineni, G., Sagaro, G. G., Chinatalapudi, N., &Amenta, F. (2020). Applications of machine learning predictive models in the chronic disease diagnosis. Journal of personalized medicine, 10(2), 21. Retrieved from https://www.mdpi.com/2075-4426/10/2/21/pdf
Chaurasia, V., & Pal, S. (2020). Application of machine learning time series analysis for prediction COVID-19 pandemic. Research on Biomedical Engineering, 1-13. Retrieved fromhttps://link.springer.com/article/10.1007/s42600-020-00105-4
Chen, M., Hao, Y., Hwang, K., Wang, L., & Wang, L. (2017). Disease prediction by machine learning over big data from healthcare communities. Ieee Access, 5, 8869-8879. Retrieved from https://ieeexplore.ieee.org/iel7/6287639/6514899/07912315.pdf
Ekum, M., &Ogunsanya, A. (2020). Application of hierarchical polynomial regression models to predict transmission of COVID-19 at global level. Int J Clin BiostatBiom, 6(1), 027. Retrieved from: https://www.researchgate.net/profile/Adeyinka-Ogunsanya/publication/342872490_Application_of_Hierarchical_Polynomial_Regression_Models_to_Predict_Transmission_of_COVID-19_at_Global_Level/links/5f0a220c92851c52d6297c0b/Application-of-Hierarchical-Polynomial-Regression-Models-to-Predict-Transmission-of-COVID-19-at-Global-Level.pdf
Ghoshal, B.; Tucker, A. Estimating Uncertainty and Interpretability in Deep Learning for Coronavirus (COVID-19) Detection. arXiv 2020, arXiv:2003.10769v3. Retrieved from https://arxiv.org/pdf/2003.10769
Gupta, A. K., Singh, V., Mathur, P., &Travieso-Gonzalez, C. M. (2021). Prediction of COVID-19 pandemic measuring criteria using support vector machine, prophet and linear regression models in Indian scenario. Journal of Interdisciplinary Mathematics, 24(1), 89-108. Retrieved from https://accedacris.ulpgc.es/bitstream/10553/76419/1/PredictionofCOVID19pandemicmeasuringcriteriausingsupportvectormachineprophetandlinearregressionmodelsinIndianscenario.pdf
Kassania, S. H., Kassanib, P. H., Wesolowskic, M. J., Schneidera, K. A., &Detersa, R. (2021). Automatic detection of coronavirus disease (COVID-19) in X-ray and CT images: a machine learning-based approach. Biocybernetics and Biomedical Engineering, 41(3), 867-879. Retrieved from: https://arxiv.org/pdf/2004.10641
Khan, Z. A., Abbasi, U., & Kim, S. W. (2021). Machine Learning and LPWAN Based Internet of Things Applications in Healthcare Sector during COVID-19 Pandemic. Electronics, 10(14), 1615. https://www.mdpi.com/2079-9292/10/14/1615/pdf
Kwekha-Rashid, A. S., Abduljabbar, H. N., &Alhayani, B. (2021). Coronavirus disease (COVID-19) cases analysis using machine-learning applications. Applied Nanoscience, 1-13. Retrieved fromhttps://link.springer.com/article/10.1007/s13204-021-01868-7
Magdon-Ismail, M. Machine Learning the Phenomenology of COVID-19 From Early Infection Dynamics. arXiv 2020, arXiv:2003.07602.Retrieved from https://arxiv.org/pdf/2003.07602
Mbunge, E., Akinnusi, B., Fashoto, S. G., Metfula, A. S., &Mashwama, P. (2021). A critical review of emerging technologies for tackling COVID‐19 pandemic. Human behavior and emerging technologies, 3(1), 25-39. Retrieved fromhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7753602/
Melnikovas, A. (2018). Towards an explicit research methodology: Adapting research onion model for future studies. Journal of Futures Studies, 23(2), 29-44. Retrieved from https://jfsdigital.org/wp-content/uploads/2019/01/03-Melnikovas-Onion-Research-Model.pdf
Monrat, A. A., Schelén, O., & Andersson, K. (2019). A survey of blockchain from the perspectives of applications, challenges, and opportunities. IEEE Access, 7, 117134-117151.https://ieeexplore.ieee.org/iel7/6287639/8600701/08805074.pdf
Moreb, M., Mohammed, T. A., &Bayat, O. (2020). A novel software engineering approach toward using machine learning for improving the efficiency of health systems. IEEE Access, 8, 23169-23178. Retrieved fromhttps://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8974224
Namasudra, S., Dhamodharavadhani, S., &Rathipriya, R. (2021). Non-linearNonlinear neural network based forecasting model for predicting COVID-19 cases. Neural Processing Letters, 1-21. Retrieved fromhttps://link.springer.com/article/10.1007/s11063-021-10495-w
Ndiaye, B.M.; Tending, L.; Seck, D. Analysis of the COVID-19 pandemic by SIR model and machine learning techniques for forecasting. arXiv 2020, arXiv:2004.01574v1. Retrieved from https://arxiv.org/pdf/2004.01574
Patil, A., Thakre, U., Aftab, M. Z., &Warekar, S. (2021). A Comparative Analysis and Illustrative Study of Machine Learning Models for COVID-19 Prediction in India. Machine Learning, 8(08). Retrieved from https://irjet.com/archives/V8/i8/IRJET-V8I8123.pdf
Shahid, F., Zameer, A., &Muneeb, M. (2020). Predictions for COVID-19 with deep learning models of LSTM, GRU and Bi-LSTM. Chaos, Solitons & Fractals, 140, 110212.https://www.ncbi.nlm.nih.gov/pmc/articles/pmc7437542/
Xue, J., Chen, J., Hu, R., Chen, C., Zheng, C., Su, Y., & Zhu, T. (2020). Twitter discussions and emotions about the COVID-19 pandemic: Machine learning approach. Journal of medical Internet research, 22(11), e20550.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7215145/
Yeşilkanat, C. M. (2020). Spatio-temporal estimation of the daily cases of COVID-19 in worldwide using random forest machine-learning algorithm. Chaos, Solitons & Fractals, 140, 110210. Retrieved fromhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7439995/
Yoo, S. H., Geng, H., Chiu, T. L., Yu, S. K., Cho, D. C., Heo, J., … & Lee, H. (2020). Deep learning-based decision-tree classifier for COVID-19 diagnosis from chest X-ray imaging. Frontiers in medicine, 7, 427. Retrieved fromhttps://www.frontiersin.org/articles/10.3389/fmed.2020.00427/full
Kassania, S. H., Kassanib, P. H., Wesolowskic, M. J., Schneidera, K. A., & Detersa, R. (2021). Automatic detection of coronavirus disease (COVID-19) in X-ray and CT images: a machine learning based approach. Biocybernetics and Biomedical Engineering, 41(3), 867-879. Retrieved from https://arxiv.org/pdf/2004.10641