PREFMoDeL: A Systematic Review and Proposed Taxonomy of Biomolecular Features for Deep Learning

Of fundamental importance in biochemical and biomedical research is understanding a molecule's biological properties—its structure, its function(s), and its activity(ies). To this end, computational methods in Artificial Intelligence, in particular Deep Learning (DL), have been applied to further biomolecular understanding—from analysis and prediction of protein–protein and protein–ligand interactions to drug discovery and design. While choosing the most appropriate DL architecture is vitally important to accurately model the task at hand, equally important is choosing the features used as input to represent molecular properties in these DL models. Through hypothesis testing, bioinformaticians have created thousands of engineered features for biomolecules such as proteins and their ligands. Herein we present an organizational taxonomy for biomolecular features extracted from 808 articles from across the scientific literature. This objective view of biomolecular features can reduce various forms of experimental and/or investigator bias and additionally facilitate feature selection in biomolecular analysis and design tasks. The resulting dataset contains 1360 nondeduplicated features, and a sample of these features were classified by their properties, clustered, and used to suggest new features. The complete feature dataset (the Public Repository of Engineered Features for Molecular Deep Learning, PREFMoDeL) is released for collaborative sourcing on the web.

Active regression model for clinical grading of COVID-19.

Background: In the therapeutic process of COVID-19, the majority of indicators that physicians have for assisting treatment have come from clinical tests represented by proteins, metabolites, and immune levels in patients' blood. Therefore, this study constructs an individualized treatment model based on deep learning methods, aiming to realize timely intervention based on clinical test indicator data of COVID-19 patients and provide an important theoretical basis for optimizing medical resource allocation. Methods: This study collected clinical data from a total of 1,799 individuals, including 560 controls for non-respiratory infectious diseases (Negative), 681 controls for other respiratory virus infections (Other), and 558 coronavirus infections (Positive) for COVID-19. We first used the Student T-test to screen for statistically significant differences (Pvalue<0.05); we then used the Adaptive-Lasso method stepwise regression to screen the characteristic variables and filter the features with low importance; we then used analysis of covariance to calculate the correlation between variables and filter the highly correlated features; and finally, we analyzed the feature contribution and screened the best combination of features. Results: Feature engineering reduced the feature set to 13 feature combinations. The correlation coefficient between the projected results of the artificial intelligence-based individualized diagnostic model and the fitted curve of the actual values in the test group was 0.9449 which could be applied to the clinical prognosis of COVID-19. In addition, the depletion of platelets in patients with COVID-19 is an important factor affecting their severe deterioration. With the progression of COVID-19, there is a slight decrease in the total number of platelets in the patient's body, particularly as the volume of larger platelets sharply decreases. The importance of plateletCV (count*mean platelet volume) in evaluating the severity of COVID-19 patients is higher than the count of platelets and mean platelet volume. Conclusion: In general, we found that for patients with COVID-19, the increase in mean platelet volume was a predictor for SARS-Cov-2. The rapid decrease of platelet volume and the decrease of total platelet volume are dangerous signals for the aggravation of SARS-Cov-2 infection. The analysis and modeling results of this study provide a new perspective for individualized accurate diagnosis and treatment of clinical COVID-19 patients.

Prediction and interpretation of daily NFT and DeFi prices dynamics: Inspection through ensemble machine learning & XAI

Non Fungible Tokens (NFT) and Decentralized Finance (DeFi) assets have seen a growing media coverage and garnered considerable investor traction despite being classified as a niche in the digital financial sector. The lack of substantial research to demystify the dynamics of NFT and DeFi coins motivates the scrupulous analysis of the said sector. This work aims to critically delve into the evolutionary pattern of the NFTs and DeFis for performing predictive analytics of the same during the COVID-19 regime. The multivariate framework comprises the systematic inclusion of explanatory features embodying technical indicators, key macroeconomic indicators, and constructs linked to media hype and sentiment pertinent to the pandemic, nonlinear feature engineering, and ensemble machine learning. Isometric Mapping (ISOMAP) and Uniform Manifold Approximation and Projection (UMAP) techniques are conjugated with Gradient Boosting Regression (GBR) and Random Forest (RF) for enabling the predictive analysis. The predictive performance rationalizes the frameworks' capacity to accurately predict the prices of the majority of the NFT and DeFi coins during the ongoing financial distress period. Additionally, Explainable Artificial Intelligence (XAI) methodologies are used to comprehend the nature of the impact of the explanatory variables. Findings suggest that the daily movement of the NFTs and DeFi highly depends on their past historical movement. © 2023 The Authors

Smart Short-Term Load Forecasting through Coordination of LSTM-Based Models and Feature Engineering Methods during the COVID-19 Pandemic

Short-term load forecasting is essential for power companies because it is necessary to ensure sufficient capacity. This article proposes a smart load forecasting scheme to forecast the short-term load for an actual sample network in the presence of uncertainties such as weather and the COVID-19 epidemic. The studied electric load data with hourly resolution from the beginning of 2020 to the first seven days of 2021 for the New York Independent Operator is the basis for the modeling. The new components used in this article include the coordination of stacked long short-term memory-based models and feature engineering methods. Also, more accurate and realistic modeling of the problem has been implemented according to the existing conditions through COVID-19 epidemic data. The influential variables for short-term load forecasting through various feature engineering methods have contributed to the problem. The achievements of this research include increasing the accuracy and speed of short-term electric load forecasting, reducing the probability of overfitting during model training, and providing an analytical comparison between different feature engineering methods. Through an analytical comparison between different feature engineering methods, the findings of this article show an increase in the accuracy and speed of short-term load forecasting. The results indicate that combining the stacked long short-term memory model and feature engineering methods based on extra-trees and principal component analysis performs well. The RMSE index for day-ahead load forecasting in the best engineering method for the proposed stacked long short-term memory model is 0.1071. © 2023 Taylor & Francis Group, LLC.

Fake news detection: deep semantic representation with enhanced feature engineering.

Due to the widespread use of social media, people are exposed to fake news and misinformation. Spreading fake news has adverse effects on both the general public and governments. This issue motivated researchers to utilize advanced natural language processing concepts to detect such misinformation in social media. Despite the recent research studies that only focused on semantic features extracted by deep contextualized text representation models, we aim to show that content-based feature engineering can enhance the semantic models in a complex task like fake news detection. These features can provide valuable information from different aspects of input texts and assist our neural classifier in detecting fake and real news more accurately than using semantic features. To substantiate the effectiveness of feature engineering besides semantic features, we proposed a deep neural architecture in which three parallel convolutional neural network (CNN) layers extract semantic features from contextual representation vectors. Then, semantic and content-based features are fed to a fully connected layer. We evaluated our model on an English dataset about the COVID-19 pandemic and a domain-independent Persian fake news dataset (TAJ). Our experiments on the English COVID-19 dataset show 4.16% and 4.02% improvement in accuracy and f1-score, respectively, compared to the baseline model, which does not benefit from the content-based features. We also achieved 2.01% and 0.69% improvement in accuracy and f1-score, respectively, compared to the state-of-the-art results reported by Shifath et al. (A transformer based approach for fighting covid-19 fake news, arXiv preprint arXiv:2101.12027, 2021). Our model outperformed the baseline on the TAJ dataset by improving accuracy and f1-score metrics by 1.89% and 1.74%, respectively. The model also shows 2.13% and 1.6% improvement in accuracy and f1-score, respectively, compared to the state-of-the-art model proposed by Samadi et al. (ACM Trans Asian Low-Resour Lang Inf Process, https://doi.org/10.1145/3472620, 2021).

Temporal analysis and opinion dynamics of COVID-19 vaccination tweets using diverse feature engineering techniques.

The outbreak of the COVID-19 pandemic has also triggered a tsunami of news, instructions, and precautionary measures related to the disease on social media platforms. Despite the considerable support on social media, a large number of fake propaganda and conspiracies are also circulated. People also reacted to COVID-19 vaccination on social media and expressed their opinions, perceptions, and conceptions. The present research work aims to explore the opinion dynamics of the general public about COVID-19 vaccination to help the administration authorities to devise policies to increase vaccination acceptance. For this purpose, a framework is proposed to perform sentiment analysis of COVID-19 vaccination-related tweets. The influence of term frequency-inverse document frequency, bag of words (BoW), Word2Vec, and combination of TF-IDF and BoW are explored with classifiers including random forest, gradient boosting machine, extra tree classifier (ETC), logistic regression, Naïve Bayes, stochastic gradient descent, multilayer perceptron, convolutional neural network (CNN), bidirectional encoder representations from transformers (BERT), long short-term memory (LSTM), and recurrent neural network (RNN). Results reveal that ETC outperforms using BoW with a 92% of accuracy and is the most suitable approach for sentiment analysis of COVID-19-related tweets. Opinion dynamics show that sentiments in favor of vaccination have increased over time.

Using Machine Learning to Predict Business Survival in COVID-19

The global recession due to the pandemic has knocked the business landscape and brought the world to its knees. There were a number of renowned companies that made the headlines for being the top industry hard hits. Nonetheless, there were businesses that survived this pandemic and navigated the COVID complexities so effectively that it tipped the scales in their favor. We attempt to study the factors that helped these businesses masterfully work their way through the conundrums of coronavirus pandemic. We first build a dataset that entailed information pertinent to businesses and relevant COVID-related information that was sourced from Yelp and other platforms. We used a variety of classifiers to make predictions about the survival of these businesses followed by that after assessing their performance through varied methods. The model efficiency was classified based on several rating techniques to evaluate both underperforming and profitable businesses. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Prediction and interpretation of daily NFT and DeFi prices dynamics: Inspection through ensemble machine learning & XAI

Non Fungible Tokens (NFT) and Decentralized Finance (DeFi) assets have seen a growing media coverage and garnered considerable investor traction despite being classified as a niche in the digital financial sector. The lack of substantial research to demystify the dynamics of NFT and DeFi coins motivates the scrupulous analysis of the said sector. This work aims to critically delve into the evolutionary pattern of the NFTs and DeFis for performing predictive analytics of the same during the COVID-19 regime. The multivariate framework comprises the systematic inclusion of explanatory features embodying technical indicators, key macroeconomic indicators, and constructs linked to media hype and sentiment pertinent to the pandemic, nonlinear feature engineering, and ensemble machine learning. Isometric Mapping (ISOMAP) and Uniform Manifold Approximation and Projection (UMAP) techniques are conjugated with Gradient Boosting Regression (GBR) and Random Forest (RF) for enabling the predictive analysis. The predictive performance rationalizes the frameworks' capacity to accurately predict the prices of the majority of the NFT and DeFi coins during the ongoing financial distress period. Additionally, Explainable Artificial Intelligence (XAI) methodologies are used to comprehend the nature of the impact of the explanatory variables. Findings suggest that the daily movement of the NFTs and DeFi highly depends on their past historical movement.

Smart Short-Term Load Forecasting through Coordination of LSTM-Based Models and Feature Engineering Methods during the COVID-19 Pandemic

Short-term load forecasting is essential for power companies because it is necessary to ensure sufficient capacity. This article proposes a smart load forecasting scheme to forecast the short-term load for an actual sample network in the presence of uncertainties such as weather and the COVID-19 epidemic. The studied electric load data with hourly resolution from the beginning of 2020 to the first seven days of 2021 for the New York Independent Operator is the basis for the modeling. The new components used in this article include the coordination of stacked long short-term memory-based models and feature engineering methods. Also, more accurate and realistic modeling of the problem has been implemented according to the existing conditions through COVID-19 epidemic data. The influential variables for short-term load forecasting through various feature engineering methods have contributed to the problem. The achievements of this research include increasing the accuracy and speed of short-term electric load forecasting, reducing the probability of overfitting during model training, and providing an analytical comparison between different feature engineering methods. Through an analytical comparison between different feature engineering methods, the findings of this article show an increase in the accuracy and speed of short-term load forecasting. The results indicate that combining the stacked long short-term memory model and feature engineering methods based on extra-trees and principal component analysis performs well. The RMSE index for day-ahead load forecasting in the best engineering method for the proposed stacked long short-term memory model is 0.1071. [ FROM AUTHOR]

Feature Selection Mechanism for Attention Classification using Gaze Tracking Data

The Covid-19 outbreak has caused disruptions in the education sector, making remote education the dominant mode for lecture delivery. The lack of visual feedback and physical interaction makes it very hard for teachers to measure the engagement level of students during lectures. This paper proposes a time-bounded window operation to extract statistical features from raw gaze data, captured in a remote teaching experiment and link them with the student's attention level. Feature selection or dimensionality reduction is performed to reduce the convergence time and overcome the problem of over-fitting. Recursive feature elimination (RFE) and SelectFromModel (SFM) are used with different machine learning (ML) algorithms, and a subset of optimal feature space is obtained based on the feature scores. The model trained using the optimal feature subset showed significant improvement in accuracy and computational complexity. For instance, a support vector classifier (SVC) led 2.39% improvement in accuracy along with approximately 66% reduction in convergence time. © 2022 IEEE.

Using Machine Learning to Predict Business Survival in COVID-19

The global recession due to the pandemic has knocked the business landscape and brought the world to its knees. There were a number of renowned companies that made the headlines for being the top industry hard hits. Nonetheless, there were businesses that survived this pandemic and navigated the COVID complexities so effectively that it tipped the scales in their favor. We attempt to study the factors that helped these businesses masterfully work their way through the conundrums of coronavirus pandemic. We first build a dataset that entailed information pertinent to businesses and relevant COVID-related information that was sourced from Yelp and other platforms. We used a variety of classifiers to make predictions about the survival of these businesses followed by that after assessing their performance through varied methods. The model efficiency was classified based on several rating techniques to evaluate both underperforming and profitable businesses. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Breast Cancer Treatment at Home During COVID Pandemic Using IoT, Cloud Computing and Machine Learning

It is crucial that Breast Cancer should be detected early. Breast cancer time series forecasting is a novel data - driven approach to breast cancer diagnosis. Instead of looking at static images of the medical records, it analyses the dynamics in the tumour's growth rate, especially in its early stages. It uses machine learning models to find patterns that are not readily observable in static images, but are predictive of later outcomes. During COVID-19 it is necessary to monitor patient from home and IOT devices can be used that give moment forecast to the client and doctor during their typical day by day routine. Various Machine learning models are reviewed for classification of Breast Cancer symptoms. It is observed that data visualization and feature engineering play a crucial role in the classification before applying any model on data set. For human protection during COVID-19 it is better to depend on IoT enabled wearable device for automatic detection and appointment. The IoT enabled devices can use power of cloud computing and machine learning models to complete the framework of getting treated at home. Security of the data is another aspect to be taken into consideration. Solutions are available for the whole process and their aggregation will result in generating the desired model. In this paper, model is proposed to diagnose breast cancer at home using IoT, Blockchain, Machine learning and Cloud Computing. © 2022 American Institute of Physics Inc.. All rights reserved.

Classification of audio signals using spectrogram surfaces and extrinsic distortion measures.

Representation of one-dimensional (1D) signals as surfaces and higher-dimensional manifolds reveals geometric structures that can enhance assessment of signal similarity and classification of large sets of signals. Motivated by this observation, we propose a novel robust algorithm for extraction of geometric features, by mapping the obtained geometric objects into a reference domain. This yields a set of highly descriptive features that are instrumental in feature engineering and in analysis of 1D signals. Two examples illustrate applications of our approach to well-structured audio signals: Lung sounds were chosen because of the interest in respiratory pathologies caused by the coronavirus and environmental conditions; accent detection was selected as a challenging speech analysis problem. Our approach outperformed baseline models under all measured metrics. It can be further extended by considering higher-dimensional distortion measures. We provide access to the code for those who are interested in other applications and different setups (Code: https://github.com/jeremy-levy/Classification-of-audio-signals-using-spectrogram-surfaces-and-extrinsic-distortion-measures). Supplementary Information: The online version contains supplementary material available at 10.1186/s13634-022-00933-9.

A New Feature Based Deep Attention Sales Forecasting Model for Enterprise Sustainable Development

In recent years, with the rise of the Internet, e-commerce has become an important field of commodity sales. However, e-commerce is affected by many factors, and the wrong judgment of supply and marketing relationships will bring huge losses to operators. Therefore, it is of great significance to establish a model that can effectively achieve high precision sales prediction for ensuring the sustainable development of e-commerce enterprises. In this paper, we propose an e-commerce sales forecasting model that considers the features of many aspects of correlation. In the first layer of the model, the temporal convolutional network (TCN) is used to extract the deep temporal characteristics of univariate sales historical data, which ensures the integrity of temporal information of sales characteristics. In the second layer, the feature selection method based on reinforcement learning is used to filter the effective correlation feature set and combine it with the temporal feature after processing, which not only improves the amount of effective information input by the model, but also avoids the high feature dimension. The third layer of the reformer model learns all the features and pays different attention to the features with different degrees of importance, ensuring the stability of the sales forecast. In the experimental part, we compare the proposed model with the current advanced sales forecasting model, and we can find that the proposed model has higher stability and accuracy.

Feature selection by machine learning models to identify the public’s changing priorities during the COVID-19 pandemic

People around the world have experienced fundamental transformations during mass events. The Industrial Revolution, World War II, and the collapse of the Berlin Wall are some of the cases that have caused radical societal changes. COVID-19 has also been a process of mass experiences regarding society. Determining the mass impact the pandemic has had on society shows that the pandemic is facilitating the transition to the so-called new normal. Istanbul is a multi-identity city where 16 million people have intensely experienced the pandemic’s impact. While determining the identities of cities in the world, one can see that different city structures provide different data sets. This study models a machine learning algorithm suitable for the data set we’ve determined for the 39 different districts of Istanbul and 82 different features of Istanbul. The aim of the study is to indicate the changing societal trends during the COVID-19 pandemic using machine learning techniques. Thus, this work contributes to the literature and real life in terms of redesigning cities for the post-COVID19 period. Another contribution of this study is that the proposed methodology provides clues on what people in cities consider important during a pandemic.

Predicting student performance using sequence classification with time-based windows.

A growing number of universities worldwide use various forms of online and blended learning as part of their academic curricula. Furthermore, the recent changes caused by the COVID-19 pandemic have led to a drastic increase in importance and ubiquity of online education. Among the major advantages of e-learning is not only improving students' learning experience and widening their educational prospects, but also an opportunity to gain insights into students' learning processes with learning analytics. This study contributes to the topic of improving and understanding e-learning processes in the following ways. First, we demonstrate that accurate predictive models can be built based on sequential patterns derived from students' behavioral data, which are able to identify underperforming students early in the course. Second, we investigate the specificity-generalizability trade-off in building such predictive models by investigating whether predictive models should be built for every course individually based on course-specific sequential patterns, or across several courses based on more general behavioral patterns. Finally, we present a methodology for capturing temporal aspects in behavioral data and analyze its influence on the predictive performance of the models. The results of our improved sequence classification technique are capable to predict student performance with high levels of accuracy, reaching 90% for course-specific models.

CoviFacts—A COVID-19 Fake News Detector Using Natural Language Processing

Fake news confronts us on a daily basis in today’s fast-paced social media world. While some instances of fake news might seem innocuous, there are many examples that prove to be menacing. Misinformation or disinformation which takes the form of these weaponized lies which eventually amount to defective information, defamatory allegations, and hoaxes. The only motive behind such a malicious act is to engender emotional instability among the public. One such prevalent example today is COVID-19 which has caused an unprecedented paradigm shift in numerous businesses and quotidian activities across the globe. One of the primary activities is being news reporting. On average, people are spending almost one hour a day reading news via many different sources. The development in technology has obviated the barriers between sharing of information, thereby truly making the industry cosmopolitan. Therefore, it is paramount to curb fake news at source and prevent it from spreading to a larger audience. This paper describes a system, where the user can identify apocryphal news related to COVID-19 so as to ensure its authenticity. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Deep learning based cough detection camera using enhanced features.

Coughing is a typical symptom of COVID-19. To detect and localize coughing sounds remotely, a convolutional neural network (CNN) based deep learning model was developed in this work and integrated with a sound camera for the visualization of the cough sounds. The cough detection model is a binary classifier of which the input is a two second acoustic feature and the output is one of two inferences (Cough or Others). Data augmentation was performed on the collected audio files to alleviate class imbalance and reflect various background noises in practical environments. For effective featuring of the cough sound, conventional features such as spectrograms, mel-scaled spectrograms, and mel-frequency cepstral coefficients (MFCC) were reinforced by utilizing their velocity (V) and acceleration (A) maps in this work. VGGNet, GoogLeNet, and ResNet were simplified to binary classifiers, and were named V-net, G-net, and R-net, respectively. To find the best combination of features and networks, training was performed for a total of 39 cases and the performance was confirmed using the test F1 score. Finally, a test F1 score of 91.9% (test accuracy of 97.2%) was achieved from G-net with the MFCC-V-A feature (named Spectroflow), an acoustic feature effective for use in cough detection. The trained cough detection model was integrated with a sound camera (i.e., one that visualizes sound sources using a beamforming microphone array). In a pilot test, the cough detection camera detected coughing sounds with an F1 score of 90.0% (accuracy of 96.0%), and the cough location in the camera image was tracked in real time.

Feature Importance Analysis by Nowcasting Perspective to Predict COVID-19

The present work raises an investigation about prediction and the feature importance to estimate the COVID-19 infection, using Machine Learning approach. Our work analyzed the inclusion of climatic features, mobility, government actions and the number of cases per health sub-territory from an existing model. The Random Forest with Permutation Importance method was used to assess the importance and list the thirty most relevant that represent the probability of infection of the disease. Among all features, the most important were: i) the variables per region health stand out, ii) period comprised between the date of notification and symptom onset, iii) symptoms features as fever, cough and sore throat, iv) variables of the traffic flow and mobility, and also v) wheathers features. The model was validated and reached an accuracy average of 81.82%, whereas the sensitivity and specificity achieved 87.52% and the 78.67% respectively in the infection estimate. Therefore, the proposed investigation represents an alternative to guide authorities in understanding aspects related to the disease. [ FROM AUTHOR] Copyright of Mobile Networks & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Prediction of psychiatric drugs sale during COVID-19

In the pharmaceutical industry, the production of psychiatric drugs has been seriously disrupted since the appearance of COVID'19. For that, Demand Forecasting of psychiatric drugs is among the big challenges in this industry. The objective is to avoid an excess of stock and, at the same time, to ensure that a stock rupture does not occur. Based on analysis of psychiatric drugs data, we compare in this paper several forecasting techniques which are Exponential Smoothing, seasonal ARIMA (i.e. SARIMA), SARIMAX, enhanced with the integration of exogenous (explanatory) variables, and LSTM. Through all the done tests, we make a comparison study of the results to identify the most promising models.