HuDA_COVID Human Disposition Analysis During COVID-19 Using Machine Learning

Coronavirus has greatly impacted various aspects of human life, including human psychology and human disposition. In this paper, the authors analyzed the impact of the COVID-19 pandemic on human health. In the proposed work, human disposition analysis during COVID-19 using machine learning (HuDA_COVID), where factors such as age, employment, addiction, stress level are studied. A mass survey is conducted on individuals of various age groups, regions, and professions, and the methodology achieved varied accuracy ranges from 87.5% to 98%. The study shows people are worried about lockdown, work, and relationships. Furthermore, 23% of the respondents have not had any effect. Forty-five percent and 32% have had positive and negative effects, respectively. HuDA_COVID is a novel study in human disposition analysis in COVID-19 where a weighted assignment indicating the health status is also proposed. HuDA_COVID clearly indicates a need for a methodical approach towards the human psychological needs to help the social organizations formulating holistic interventions for affected individuals.

Robust Classification and Detection of Big Medical Data Using Advanced Parallel K-Means Clustering, YOLOv4, and Logistic Regression.

Big-medical-data classification and image detection are crucial tasks in the field of healthcare, as they can assist with diagnosis, treatment planning, and disease monitoring. Logistic regression and YOLOv4 are popular algorithms that can be used for these tasks. However, these techniques have limitations and performance issue with big medical data. In this study, we presented a robust approach for big-medical-data classification and image detection using logistic regression and YOLOv4, respectively. To improve the performance of these algorithms, we proposed the use of advanced parallel k-means pre-processing, a clustering technique that identified patterns and structures in the data. Additionally, we leveraged the acceleration capabilities of a neural engine processor to further enhance the speed and efficiency of our approach. We evaluated our approach on several large medical datasets and showed that it could accurately classify large amounts of medical data and detect medical images. Our results demonstrated that the combination of advanced parallel k-means pre-processing, and the neural engine processor resulted in a significant improvement in the performance of logistic regression and YOLOv4, making them more reliable for use in medical applications. This new approach offers a promising solution for medical data classification and image detection and may have significant implications for the field of healthcare.

DT2CAM: A ecision ree to ontent ddressable emory Framework

Decision trees are powerful tools for data classification. Accelerating the decision tree search is crucial for on-the-edge applications with limited power and latency budget. In this paper, we propose a content-addressable memory compiler for decision tree inference acceleration. We propose a novel ”adaptive-precision”scheme that results in a compact implementation and enables an efficient bijective mapping to ternary content addressable memories while maintaining high inference accuracies. We also develop a resistive-based functional synthesizer to map the decision tree to resistive content addressable memory arrays and perform functional simulations for energy, latency, and accuracy evaluations. We study the decision tree accuracy under hardware non-idealities including device defects, manufacturing variability, and input encoding noise. We test our framework on various decision tree datasets including Give Me Some Credit, Titanic, and COVID-19. Our results reveal up to 42.4%energy savings and up to <inline-formula><tex-math notation="LaTeX">$17.8\times$</tex-math></inline-formula> better energy-delay-area product compared to the state-of-art hardware accelerators, and up to 333 million decisions per sec for the pipelined implementation. IEEE

A Hybrid Approach for Detection of Fake News in Sinhala Text

The impact of technology on people's lives has grown continuously. The consumption of online news is one of the important trends as the share of population with internet access grows rapidly over time. Global statistics have shown that the internet and social media usage has an increasing trend. Recent developments like the Covid 19 pandemic have amplified this trend even more. However, the credibility of online news is a very critical issue to consider since it directly impacts the society and the people's mindsets. Majority of users tend to instinctively believe what they encounter and come into conclusions based upon them. It is essential that the consumers have an understanding or prior knowledge regarding the news and its source before coming into conclusions. This research proposes a hybrid model to predict the accuracy of a particular news article in Sinhala text. The model combines the general news content based analysis techniques using machine learning/ deep learning classifiers with social network related features of the news source to make predictions. A scoring mechanism is utilized to provide an overall score to a given news item where two independent scores- Accuracy Score (by analyzing the news content) and Credibility Score (by a scoring mechanism on social network features of the news source) are combined. The hybrid model containing the Passive Aggressive Classifier has shown the highest accuracy of 88%. Also, the models containing deep neural netWorks has shown accuracy around 75-80%. These results highlight that the proposed method could efficiently serve as a Fake News Detection mechanism for news content in Sinhala Language. Also, since there's no publicly available dataset for Fake News detection in Sinhala, the datasets produced in this work could also be considered as a contribution from this research. © 2022 IEEE.

Characterizing the Differences in Descriptions of Violence on Reddit During the COVID-19 Pandemic.

Concerns have been raised over the experiences of violence such as domestic violence (DV) and intimate partner violence (IPV) during the COVID-19 pandemic. Social media such as Reddit represent an alternative outlet for reporting experiences of violence where healthcare access has been limited. This study analyzed seven violence-related subreddits to investigate the trends of different violence patterns from January 2018 to February 2022 to enhance the health-service providers' existing service or provide some new perspective for existing violence research. Specifically, we collected violence-related texts from Reddit using keyword searching and identified six major types with supervised machine learning classifiers: DV, IPV, physical violence, sexual violence, emotional violence, and nonspecific violence or others. The increase rate (IR) of each violence type was calculated and temporally compared in five phases of the pandemic. The phases include one pre-pandemic phase (Phase 0, the date before February 26, 2020) and four pandemic phases (Phases 1-4) with separation dates of June 17, 2020, September 7, 2020, and June 4, 2021. We found that the number of IPV-related posts increased most in the earliest phase; however, that for COVID-citing IPV was highest in the mid-pandemic phase. IRs for DV, IPV, and emotional violence also showed increases across all pandemic phases, with IRs of 26.9%, 58.8%, and 28.8%, respectively, from the pre-pandemic to the first pandemic phase. In the other three pandemic phases, all the IRs for these three types of violence were positive, though lower than the IRs in the first pandemic phase. The findings highlight the importance of identifying and providing help to those who suffer from such violent experiences and support the role of social media site monitoring as a means of informative surveillance for help-providing authorities and violence research groups.

Characterizing data patterns with core–periphery network modeling

Traditional classification techniques usually classify data samples according to the physical organization, such as similarity, distance, and distribution, of the data features, which lack a general and explicit mechanism to represent data classes with semantic data patterns. Therefore, the incorporation of data pattern formation in classification is still a challenge problem. Meanwhile, data classification techniques can only work well when data features present high level of similarity in the feature space within each class. Such a hypothesis is not always satisfied, since, in real-world applications, we frequently encounter the following situation: On one hand, the data samples of some classes (usually representing the normal cases) present well defined patterns;on the other hand, the data features of other classes (usually representing abnormal classes) present large variance, i.e., low similarity within each class. Such a situation makes data classification a difficult task. In this paper, we present a novel solution to deal with the above mentioned problems based on the mesostructure of a complex network, built from the original data set. Specifically, we construct a core–periphery network from the training data set in such way that the normal class is represented by the core sub-network and the abnormal class is characterized by the peripheral sub-network. The testing data sample is classified to the core class if it gets a high coreness value;otherwise, it is classified to the periphery class. The proposed method is tested on an artificial data set and then applied to classify x-ray images for COVID-19 diagnosis, which presents high classification precision. In this way, we introduce a novel method to describe data pattern of the data "without pattern” through a network approach, contributing to the general solution of classification. © 2022 Elsevier B.V.

Use of Classification Techniques for the Analysis of Data Related to COVID-19 in México

SARS-CoV-2 has bought many challenges to the world, socially, economically, and healthy habits. Even to those that have not experienced the sickness itself, and even though it has changed the lifestyle of the people across the world nation wise the effects of COVID-19 need to be analyzed and understood, analyzing a large amount of data is a process by itself, in this document details the analysis of the data collected from México by the Secretary of Health, the data was analyzed by implementing statistics, and classification methods known as K-Means, C&R Tree and TwoStep Cluster, using processed and unprocessed data. With the main emphasis on K-means. The study has the purpose of detecting what makes the highest impact on a person, to get sick, and succumb to the effects of the disease. In the study, it was found that in México the age of risk is at its highest at the age of 57, and the ones at the highest risk of mortality are those with hypertension and obesity, with those that present both at the age of 57 having a 19.37% of death. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Comparison of Naïve Bayes, C4.5 and K-Nearest Neighbor for Covid-19 Data Classification

Covid-19 is a new virus that appeared in the city of Wuhan in 2019. This virus spreads very quickly even to Indonesia. One effort that can be done to detect the presence of this virus is the PCR and antigen test. Increasing this case resulted in a medical team having difficulty detecting suspects exposed to viruses. This research was conducted to find the best classification algorithm in predicting and classifying status on the suspected Covid-19 both exposed or not exposed. The method used in this study is Naïve Bayes, C4.5 and K-Nearest Neighbor which have very high accuracy using secondary data from the Dumai City Health Agency. From this study it was found that the algorithm C4.5 as the best algorithm in predicting the status of COVID-19 patients, especially in the city of Dumai with an accuracy of 86.54%, recall 71.51%and precision 85.14%. This study has implications for further researchers in choosing an algorithm to predict the COVID-19 case. © 2022 IEEE.

Workflow for Medical Data Classification and Analysis

An approach for automated knowledge extraction and decision-making from medical images through a workflow for preprocessing of incoming X-ray images, analysis, classification and evaluation of the results is presented in this paper. The designed algorithm for analysis of medical X-rays images is based on machine learning and consists of three main phases: preprocessing of training and validation datasets, medical images classification utilizing Logistic Regression, Naive Bayes, SVM methods, evaluation of the model. A workflow was developed to process and analyze datasets of lung X-ray images containing four classes, and determine classification accuracy by examining performance evaluation parameters. © 2022 IEEE.

A Smart Prevention Management in Gestational Diabetes Mellitus

Gestational Diabetes Mellitus (GDM) is a high blood glucose level during pregnancy. Patients have frequent follow-ups throughout pregnancy. The concept of health technology enables the accessibility and efficiency of therapy in terms of time-saving and promotes adherence, especially during Covid-19. The study conducts the predictive of GDM risk using a data classification model, which has high accuracy (more than 90%). The model is used for improving the patient's self-awareness through the color notification feature. In addition, we design the GDM's system, including the electronic health information exchange, to ensure interoperability, improve service accessibility and increase patient participation. Finally, this prototype is evaluated by medical staff using the Technology Acceptance Model. The results are satisfactory and accepted because the data technology and standard are incorporated to deliver high performance. Besides, this system is expected to reduce workload and provide convenience. © 2022 IEEE.

Characterizing data patterns with core-periphery network modeling

Traditional classification techniques usually classify data samples according to the physical organization, such as similarity, distance, and distribution, of the data features, which lack a general and explicit mechanism to represent data classes with semantic data patterns. Therefore, the incorporation of data pattern formation in classification is still a challenge problem. Meanwhile, data classification techniques can only work well when data features present high level of similarity in the feature space within each class. Such a hypothesis is not always satisfied, since, in real-world applications, we frequently encounter the following situation: On one hand, the data samples of some classes (usually representing the normal cases) present well defined patterns;on the other hand, the data features of other classes (usually representing abnormal classes) present large variance, i.e., low similarity within each class. Such a situation makes data classification a difficult task. In this paper, we present a novel solution to deal with the above mentioned problems based on the mesostructure of a complex network, built from the original data set. Specifically, we construct a core–periphery network from the training data set in such way that the normal class is represented by the core sub-network and the abnormal class is characterized by the peripheral sub-network. The testing data sample is classified to the core class if it gets a high coreness value;otherwise, it is classified to the periphery class. The proposed method is tested on an artificial data set and then applied to classify x-ray images for COVID-19 diagnosis, which presents high classification precision. In this way, we introduce a novel method to describe data pattern of the data “without pattern” through a network approach, contributing to the general solution of classification.

Hybrid Mathematical Model for Data Classification and Prediction: Case study COVID-19

The increase of data availability has stimulated researchers to benefit from this data in predicting the hidden pattern for knowledge discovery. Data classification and machine learning algorithms are becoming important tools used in knowledge discovery. In this paper, we propose a hybrid classification model that combines some features and parameters from a probabilistic model and some other parameters from a divide and conquer model in a linear one. In our model, we generate a set of functions related to the number of attributes and the value of each attribute. Afterwards, these functions are reduced according to the number of classes needed. We test our model on collected data about symptoms in people infected with COVID-19 in England. Our simulation results show an accuracy rate in the range 50-80%. We expect to increase the accuracy rate if we increase the size of data used or we increase the number of attributes. © 2022. All Rights Reserved.

A functional data analysis approach to assess the prognostic value of SARS-CoV-2 infections surrogate data

COVID-19 is characterised by quite diverse prognosis. While the majority of infected individuals present no or very mild symptoms, some individuals develop severe disease requiring intensive care. This work leverages the parameters of a virtual cohort of infected individuals generated by a computational immunology model. In so doing we identify the most relevant immunological parameters for the classification of severe COVID-19 cases. The functional data analysis approach used turns out to be appropriate to analyse the output of the computational model. In this work, we classify the disease prognosis using both statistical models and machine learning algorithms adapted from functional data analysis and we compare their performances. © 2021 IEEE.

Classification of Dispersed Patterns of Radiographic Images with COVID-19 by Core-Periphery Network Modeling

In real world data classification tasks, we always face the situations where the data samples of the normal cases present a well defined pattern and the features of abnormal data samples vary from one to another, i.e., do not show a regular pattern. Up to now, the general data classification hypothesis requires the data features within each class to present a certain level of similarity. Therefore, such real situations violate the classic classification condition and make it a hard task. In this paper, we present a novel solution for this kind of problems through a network approach. Specifically, we construct a core-periphery network from the training data set in such way that core node set is formed by the normal data samples and peripheral node set contains the abnormal samples of the training data set. The classification is made by checking the coreness of the testing data samples. The proposed method is applied to classify radiographic image for COVID-19 diagnosis. Computer simulations show promising results of the method. The main contribution is to introduce a general scheme to characterize pattern formation of the data “without pattern”. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

A hybrid feature selection model based on butterfly optimization algorithm: COVID-19 as a case study.

The need to evolve a novel feature selection (FS) approach was motivated by the persistence necessary for a robust FS system, the time-consuming exhaustive search in traditional methods, and the favourable swarming manner in various optimization techniques. Most of the datasets have a high dimension in many issues since all features are not crucial to the problem, which reduces the algorithm's accuracy and efficiency. This article presents a hybrid feature selection approach to solve the low precision and tardy convergence of the butterfly optimization algorithm (BOA). The proposed method is dependent on combining the algorithm of BOA and the particle swarm optimization (PSO) as a search methodology using a wrapper framework. BOA is started with a one-dimensional cubic map in the proposed approach, and a non-linear parameter control technique is also implemented. To boost the basic BOA for global optimization, PSO algorithm is mixed with the butterfly optimization algorithm (BOAPSO). A 25 dataset evaluates the proposed BOAPSO to determine its efficiency with three metrics: classification precision, the selected features, and the computational time. A COVID-19 dataset has been used to evaluate the proposed approach. Compared to the previous approaches, the findings show the supremacy of BOAPSO for enhancing performance precision and minimizing the number of chosen features. Concerning the accuracy, the experimental outcomes demonstrate that the proposed model converges rapidly and performs better than with the PSO, BOA, and GWO with improvement percentages: 91.07%, 87.2%, 87.8%, 87.3%, respectively. Moreover, the proposed model's average selected features are 5.7 compared to the PSO, BOA, and GWO, with average features 22.5, 18.05, and 23.1, respectively.