A Novel Machine Learning-assisted Pairs Trading Approach for Trading Risk Reduction

The cryptocurrency market has been growing rapidly in recent years. The volume of transactions and the number of participants in the cryptocurrency market makes it huge enough that we cannot ignore it. At the same time, the global stock market has also reached a new height in the past two years. However, due to the COVID epidemic and other political and economic-related factors in the last two years, the uncertainty in the capital market remains high, and short-term large fluctuations occur frequently;thus, many investors have suffered substantial losses. Pairs trading, an advanced statistical arbitrage method, is believed to hedge the risk and profit off the market regardless of market condition. Amongst the vast literature on pairs trading, there have been investors trading a pair of cryptocurrencies or a pair of stocks using machine learning or empirical methods. This research probes the boundary of utilizing machine learning methods to do pairs trading with one stock asset and another cryptocurrency. Briefly, we built an assets pool with both stocks and cryptocurrencies to find the best trading pair. In addition, we applied mainstream machine learning models to the trading strategy. We finally evaluated the accuracy of the proposed method in prediction and compared their returns based on the actual U.S. Stock and Cryptocurrency Market data. The test results show that our method outperforms other state-of-the-art methods. © 2022 IEEE.

A Survey on Yogic Posture Recognition

Yoga has been a great form of physical activity and one of the promising applications in personal health care. Several studies prove that yoga is used as one of the physical treatments for cancer, musculoskeletal disorder, depression, Parkinson's disease, and respiratory heart diseases. In yoga, the body should be mechanically aligned with some effort on the muscles, ligaments, and joints for optimal posture. Postural-based yoga increases flexibility, energy, overall brain activity and reduces stress, blood pressure, and back pain. Body Postural Alignment is a very important aspect while performing yogic asanas. Many yogic asanas including uttanasana, kurmasana, ustrasana, and dhanurasana, require bending forward or backward, and if the asanas are performed incorrectly, strain in the joints, ligaments, and backbone can result, which can cause problems with the hip joints. Hence it is vital to monitor the correct yoga poses while performing different asanas. Yoga posture prediction and automatic movement analysis are now possible because of advancements in computer vision algorithms and sensors. This research investigates a thorough analysis of yoga posture identification systems using computer vision, machine learning, and deep learning techniques.

Comparison of Different Models in Predicting COVID-19 Severity Based on Chest X-Ray Scans

The global outbreak of COVID-19 has resulted in a surge in patients in hospitals and intensive care units. This unprecedented demand for medical resources has severely burdened healthcare systems. Chest X-Ray (CXR) images can be used by hospitals and small clinics to predict COVID-19 severity to maximize efficiency and allot medical resources to patients with severe COVID-19. This research compares the accuracies of four convolutional neural network models in predicting COVID-19 severity using chest X-Rays images. The CNN models include VGG-16, ResNet 50, Xception, and a custom CNN model. Through the comparison, VGG-16 had the highest COVID-19 severity prediction accuracy of all four models, with 95.56% testing accuracy and 88.33% validation accuracy. Using a machine learning method, disease progression can be tracked more accurately and help prioritize patients to ensure effective and timely treatment. © 2022 IEEE.

Performance of Machine Learning Methods Using Tweets

Public opinions shared in common platforms like Twitter, Facebook, Instagram, etc. act as the sources of information for experts. Transportation and analysis of such data is very important and difficult due to data regulations and its structure. The pre-processing approaches and word-based dictionaries are used to understand the unprocessed data and make possible the opinions/tweets to be analyzed. Machine learning algorithms learn from past experience and use a variety of statistical, probabilistic and optimization algorithms to detect useful patterns from unstructured data sets. Our study aims to compare the performance of classification algorithms to predict individuals with COVID-19(+ ) or COVID-19(−) using the emotions among the tweets by text mining procedures. Logistic Regression (LR), Support Vector Machine (SVM), Naive Bayes (NB), Decision Trees (DT), Random Forest (RF), Artificial Neural Networks (ANN), Gradient Boost (GBM) and XGradient algorithms were used to extract the accuracy of model performance of each model for the detection and identification of the disease related to the COVID-19 virus, which has been on the agenda recently. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Covid Wave Prediction using SARIMA Machine Learning Algorithm

A speedy and accurate diagnosis of COVID-19 is made possible by effective SARS-Co V -2 screening, which can also lessen the strain on health care systems. There have been built prediction models that assess the likelihood of infection by combining a number of parameters. These are intended to help medical professionals worldwide prioritize patients, particularly when there are few healthcare resources available. From a dataset of 51,831 tested people, out of which 4,769 were confirmed to have COVID-19 virus, a machine learning method was developed and trained. Records of the following week with 47,401 tested people, of which 3,624 were tested positive was also considered. Our method accurately predicted the COVID-19 test results using eight binary characteristics, including gender, age 60, known contact with an infected person, and the presence of five early clinical signs. © 2023 IEEE.

A Prototype for Interpretation of Sars-Cov-2 Tests Using Artificial Vision

Covid-19 has been declared a pandemic by the World Health Organization in March 2020, so science has been trying to help mitigate its effects from its various fields of study. Machine learning methods can play an important role in identifying test results that reveal whether an individual has the disease. This degree work presents a prototype based on computer vision and machine learning techniques to automatically detect SARS-CoV-2 serology tests. The goal of the prototype is to identify and classify the serology test cassette result by Immunoglobulin G and Immunoglobulin M indicators that are flagged after a test reaction time which is approximately 15 minutes. The results in the identification performed by the prototype are promising and ease its analysis, reducing the errors in the identification of the test and the interpretation of the results. The result is a prototype that allows to perform, simplify and improve the tasks of health professionals, which they must perform daily in the triage area. © 2022 IEEE.

A Decision Support System using Machine Learning -IoT to tackle the Covid-19 pandemic in Uttarakhand

Covid-19 spread is worldwide;India is now at the second place where this epidemic is spreading with high rate. The state of Uttarakhand, a hilly state of India also has a significant impact of Covid-19. This paper suggests that machine learning techniques with IOT can equipped the doctors, and lab technicians to deal with this pandemic. Here, we also design a prediction system to help the doctors so that they can keep the records of infected patients. We used IoT, machine learning and ensemble methods for healthcare to store infected patient's data in the cloud database, and enable doctors/others to screen patient's data about their disease. We developed a decision support system to detect the diseases quickly and the treatment can be initiated immediately. © 2022 IEEE.

A New Hybrid Artificial Intelligence Model for Diseases Identification

Artificial intelligence is increasingly applied in many fields, specially in medicine to assist patients and physicians. Growing datasets provide a sound basis to adapt machine learning methods to identify and detect some diseases. These later, are often very similar which make difficult their identification by chest X-ray images. In this paper, we introduce a diagnostic AI model that allow to separate, diagnose and classify three various diseases: tuberculosis, covid19 and Pneumonia. The proposed model is based on a combination of Deep Learning using the deep SqueezeNet model and Machine Learning: SVM, KNN, Logistic Regression, decision tree and Naive Bayes. The model is applied to a chest X-ray dataset containing images for each type of disease. To train and test our model, we split the image dataset into two training and test subsets in order to differentiate between different disease types. The accuracy show clearly that our model provides better results of diagnosis and identification. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Modeling the Optimal Grocery Store Trading Area Using Machine Learning Methods

Over the past few years, the COVID-19 pandemic has significantly transformed consumer behavior, which has undoubtedly affected a large number of industries. Food retail was among the sectors where the effect was significant and led to the transformation of the approach to customer interaction. A large part of consumers began to use online delivery services more, and key players were able to provide delivery of products with their own delivery services or third-party on-demand courier service companies. Undoubtedly, in addition to operational changes in retailers' business model, this also affected their investment activities. Some key players began to reduce their trading floor areas to increase financial efficiency and look for options to work in a convenience store format. In our research, we offer an approach for making the right investment decisions when opening a new store to balance financial metrics and customer satisfaction indicators, which is a key sales driver for the segment of customers who substitute delivery service for brick-and-mortar store visits. Using Machine Learning methods, we solve the task of scenario modeling of revenue and operational efficiency metrics for different areas of the store's trading floor, which allows us to identify the optimal choice for the retailer. Using traffic metrics during peak operation hours, we determine the minimal density of the trading area that will not lead to a decrease in the activity of guests inside the store. Such an approach allows us to evaluate the best format of the store, forecast the object's revenue, and recommend investment project parameters. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

Evaluation of the Improvements of Transformer Prediction with Convolution

Proposed and developed 5 years ago, Transformer has been a prevailing machine learning method and is widely used to solve various kinds of practical problems [1]. According to relevant works, Transformer has performed well in both natural language processing and computer vision tasks, so we would like to test its effectiveness in prediction, specifically, time series prediction. Over the past two years, COVID-19 is no doubt one of the major factors that influences the changes in the stock prices, and the medical industry should be among the most significantly affected, which would provide an ideal sample for us to study transformer on time series prediction. In this paper, we not only construct a machine learning model using Transformer to predict the stock prices of one medical company but also add a convolution layer to try to optimize the predictions. The comparison of the outcome from the two models suggests that the convolution layer could improve the performance of the naive transformer in several ways. © 2022 IEEE.

The Relationship of Public Health with Indicators of the Road Transport System

The paper presents a comparative analysis of the transport system of Russia by 12 indicators in accordance with the incidence of respiratory organs according to Rosstat data in 2019 and 2020. Machine learning methods have been applied, namely, data analysis was carried out using 9 available classification methods collected in the Data Master Azforus (DMA) program. In this program "Autoclassing” was carried out, which runs nine available methods on the same training sample. The conducted studies have demonstrated the effectiveness of using machine learning methods to identify patterns linking the health status of the population, including respiratory morbidity, with indicators of the transport system. In the course of the work, a high statistical significance of differences between classes of regions of the Russian Federation, which differ in the dynamics of Covid-19, was obtained by the most important indicators of transport system. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

A Face Mask Detection Algorithm Based on YOLOv5

As a new machine learning method, deep learning has been widely used in computer vision. YOLOv5, a target detection algorithm based on deep learning, has a good detection effect. In the case of COVID-19, masks should be worn correctly in public places. Therefore, it is urgent to design an accurate and effective face mask detection algorithm. To solve the problem of mask-wearing detection, a face mask detection algorithm based on YOLOv5 is proposed. The main research contents include training of the YOLOv5 model, verification of face mask detection function, and analysis and comparison of detection effects of three different sizes of detection models: YOLOv5s, YOLOv5m and YOLOv5l. The proposed model realizes the mask detection function and obtains the advantages and disadvantages of different scale models through performance evaluation. The maximum mAP of the model reached 88.1%, with good detection accuracy. © 2022 IEEE.

Performance of Machine Learning Methods Using Tweets

Public opinions shared in common platforms like Twitter, Facebook, Instagram, etc. act as the sources of information for experts. Transportation and analysis of such data is very important and difficult due to data regulations and its structure. The pre-processing approaches and word-based dictionaries are used to understand the unprocessed data and make possible the opinions/tweets to be analyzed. Machine learning algorithms learn from past experience and use a variety of statistical, probabilistic and optimization algorithms to detect useful patterns from unstructured data sets. Our study aims to compare the performance of classification algorithms to predict individuals with COVID-19(+ ) or COVID-19(−) using the emotions among the tweets by text mining procedures. Logistic Regression (LR), Support Vector Machine (SVM), Naive Bayes (NB), Decision Trees (DT), Random Forest (RF), Artificial Neural Networks (ANN), Gradient Boost (GBM) and XGradient algorithms were used to extract the accuracy of model performance of each model for the detection and identification of the disease related to the COVID-19 virus, which has been on the agenda recently. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Evaluating COVID-19 Sequence Data Using Nearest-Neighbors Based Network Model

The SARS-CoV-2 coronavirus is the cause of the COVID-19 disease in humans. Like many coronaviruses, it can adapt to different hosts and evolve into different lineages. It is well-known that the major SARS-CoV-2 lineages are characterized by mutations that happen predominantly in the spike protein. Understanding the spike protein structure and how it can be perturbed is vital for understanding and determining if a lineage is of concern. These are crucial to identifying and controlling current outbreaks and preventing future pandemics. Machine learning (ML) methods are a viable solution to this effort, given the volume of available sequencing data, much of which is unaligned or even unassembled. However, such ML methods require fixed-length numerical feature vectors in Euclidean space to be applicable. Similarly, euclidean space is not considered the best choice when working with the classification and clustering tasks for biological sequences. For this purpose, we design a method that converts the protein (spike) sequences into the sequence similarity network (SSN). We can then use SSN as an input for the classical algorithms from the graph mining domain for the typical tasks such as classification and clustering to understand the data. We show that the proposed alignment-free method is able to outperform the current SOTA method in terms of clustering results. Similarly, we are able to achieve higher classification accuracy using well-known Node2Vec-based embedding compared to other baseline embedding approaches. © 2022 IEEE.

Driving Factors of Polarization on Twitter During Protests Against COVID-19 Mitigation Measures in Vienna

We conduct the analysis of the Twitter discourse related to the anti-lockdown and anti-vaccination protests during the so-called 4th wave of COVID-19 infections in Austria (particularly in Vienna). We focus on predicting users' protest activity by leveraging machine learning methods and individual driving factors such as language features of users supporting/opposing Corona protests. For evaluation of our methods we utilize novel datasets, collected from discussions about a series of protests on Twitter (40488 tweets related to 20.11.2021;7639 from 15.01.2022 – the two biggest protests as well as 192 from 22.01.2022;8412 from 11.12.2021;3945 from 11.02.2022). We clustered users via the Louvain community detection algorithm on a retweet network into pro- and anti-protest classes. We show that the number of users engaged in the discourse and the share of users classified as pro-protest are decreasing with time. We have created language-based classifiers for single tweets of the two protest sides – random forest, neural networks and a regression-based approach. To gain insights into language-related differences between clusters we also investigated variable importance for a word-list-based modeling approach. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Global analysis and prediction scenario of infectious outbreaks by recurrent dynamic model and machine learning models: A case study on COVID-19.

It is essential to evaluate patient outcomes at an early stage when dealing with a pandemic to provide optimal clinical care and resource management. Many methods have been proposed to provide a roadmap against different pandemics, including the recent pandemic disease COVID-19. Due to recurrent epidemic waves of COVID-19, which have been observed in many countries, mathematical modeling and forecasting of COVID-19 are still necessary as long as the world continues to battle against the pandemic. Modeling may aid in determining which interventions to try or predict future growth patterns. In this article, we design a combined approach for analyzing any pandemic in two separate parts. In the first part of the paper, we develop a recurrent SEIRS compartmental model to predict recurrent outbreak patterns of diseases. Due to its time-varying parameters, our model is able to reflect the dynamics of infectious diseases, and to measure the effectiveness of the restrictive measures. We discuss the stable solutions of the corresponding autonomous system with frozen parameters. We focus on the regime shifts and tipping points; then we investigate tipping phenomena due to parameter drifts in our time-varying parameters model that exhibits a bifurcation in the frozen-in case. Furthermore, we propose an optimal numerical design for estimating the system's parameters. In the second part, we introduce machine learning models to strengthen the methodology of our paper in data analysis, particularly for prediction scenarios. We use MLP, RBF, LSTM, ANFIS, and GRNN for training and evaluation of COVID-19. Then, we compare the results with the recurrent dynamical system in the fitting process and prediction scenario. We also confirm results by implementing our methods on the released data on COVID-19 by WHO for Italy, Germany, Iran, and South Africa between 1/22/2020 and 7/24/2021, when people were engaged with different variants including Alpha, Beta, Gamma, and Delta. The results of this article show that the dynamic model is adequate for long-term analysis and data fitting, as well as obtaining parameters affecting the epidemic. However, it is ineffective in providing a long-term forecast. In contrast machine learning methods effectively provide disease prediction, although they do not provide analysis such as dynamic models. Finally, some metrics, including RMSE, R-Squared, and accuracy, are used to evaluate the machine learning models. These metrics confirm that ANFIS and RBF perform better than other methods in training and testing zones.

Application of classical and advanced machine learning models to predict personality on social media

Knowing personality traits and how people tend to think, feel and behave has been always an appealing and studied topic. This interest together with the vast amount of data generated every day on social networks present an ideal scenario to address this problem. By properly processing this data, it could be useful for many aspects of people's daily life. In this study, we applied different Machine Learning methods to solve this problem using a dataset labelled with the MBTI personalities, and we compared several algorithms such as Naïve Bayes, Logistic Regression and three different Artificial Neural Networks. Two main experiments were conducted. First, a clustering-oriented solution. Second, a classification approach. The latter turned out to outperform the clustering methods. On average, our models achieved around 90% accuracy. Finally, in order to show an example of our solution, we will validate our model with the latest news about COVID-19 and the La Palma Volcano. © 2023 Elsevier Ltd

Analysis of People's Dissatisfaction on COVID-19 using Discontent Questionnaire Data

The COVID-19 outbreak and accompanying policies for prevention and control, such as lockdown and movement constraints, have influenced many areas of society and every aspect of everyone's life and work. To find out factors that have the biggest influences on people's lives, we use the BERT model to classify 'Discontent Questionnaire Data on COVID-19' and apply machine learning methods to evaluate the accuracy of the results. The results show the top three influencing factors are work, stress, and worry about the future, and the classification results show a high degree of consistency and correlation. © 2022 IEEE.

A First Attempt at Unreliable News Detection in Swedish

Throughout the COVID-19 pandemic, a parallel infodemic has also been going on such that the information has been spreading faster than the virus itself. During this time, every individual needs to access accurate news in order to take corresponding protective measures, regardless of their country of origin or the language they speak, as misinformation can cause significant loss to not only individuals but also society. In this paper we train several machine learning models (ranging from traditional machine learning to deep learning) to try to determine whether news articles come from either a reliable or an unreliable source, using just the body of the article. Moreover, we use a previously introduced corpus of news in Swedish related to the COVID-19 pandemic for the classification task. Given that our dataset is both unbalanced and small, we use subsampling and easy data augmentation (EDA) to try to solve these issues. In the end, we realize that, due to the small size of our dataset, using traditional machine learning along with data augmentation yields results that rival those of transformer models such as BERT. © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0.

Azolla Quantity Prediction using WSN and Machine Learning

Due to COVID-19 pandemic, the expenditures on pellets and feeds in broiler and fish industries increase every year, leading to price overshoots in various agricultural products. Azolla is an emerging protein source alternative for tilapia and other livestock breeders that is known for its fast reproduction. This study aims to enhance the yield production of Azolla ponds in Nevalga Farm, Brgy. Sala, City of Cabuyao, Laguna by employing wireless sensor network (WSN) technology and predictive machine-learning (ML) methods. LoRa-based WSN was designed to measure the parameters that affect the growth and reproduction of Azolla. Throughout the 24-day monitoring period, the average received signal strength indication (RSSI) and signal-to-noise ratio (SNR) of the packets from the three sensing nodes ranged from -50.86 dBm to -71.39 dBm and 8.92 dB to 9.81 dB, respectively. A total of 3582 data sets were obtained during the observation. Among the three regression ML models used, K-Nearest Neighbor algorithm outperformed Linear Regression and Support Vector Machine in predicting Azolla quantity parameters on both training and validation datasets by yielding the smallest values of root mean square error (RMSE) and absolute error on the seven quantity indicators and achieving squared correlation that varied from 0.935 to 0.997. © 2022 IEEE.