COVID-19 Spread Forecasting, Mathematical Methods vs. Machine Learning, Moscow Case

To predict the spread of the new coronavirus infection COVID-19, the critical values of spread indicators have been determined for deciding on the introduction of restrictive measures using the city of Moscow as an example. A model was developed using classical methods of mathematical modeling based on exponential regression, the accuracy of the forecast was estimated, and the shortcomings of mathematical methods for predicting the spread of infection for more than two weeks. As a solution to the problem of the accuracy of long-term forecasts for more than two weeks, two models based on machine learning methods are proposed: a recurrent neural network with two layers of long short-term memory (LSTM) blocks and a 1-D convolutional neural network with a description of the choice of an optimization algorithm. The forecast accuracy of ML models was evaluated in comparison with the exponential regression model and one another using the example of data on the number of COVID-19 cases in the city of Moscow.

Download Full-text

HARTH: A Human Activity Recognition Dataset for Machine Learning

Sensors ◽

10.3390/s21237853 ◽

2021 ◽

Vol 21 (23) ◽

pp. 7853

Author(s):

Aleksej Logacjov ◽

Kerstin Bach ◽

Atle Kongsvold ◽

Hilde Bremseth Bårdstu ◽

Paul Jarle Mork

Keyword(s):

Neural Network ◽

Machine Learning ◽

Support Vector Machine ◽

Convolutional Neural Network ◽

Activity Recognition ◽

Human Activity ◽

Short Term Memory ◽

Human Activity Recognition ◽

Support Vector ◽

Free Living

Existing accelerometer-based human activity recognition (HAR) benchmark datasets that were recorded during free living suffer from non-fixed sensor placement, the usage of only one sensor, and unreliable annotations. We make two contributions in this work. First, we present the publicly available Human Activity Recognition Trondheim dataset (HARTH). Twenty-two participants were recorded for 90 to 120 min during their regular working hours using two three-axial accelerometers, attached to the thigh and lower back, and a chest-mounted camera. Experts annotated the data independently using the camera’s video signal and achieved high inter-rater agreement (Fleiss’ Kappa =0.96). They labeled twelve activities. The second contribution of this paper is the training of seven different baseline machine learning models for HAR on our dataset. We used a support vector machine, k-nearest neighbor, random forest, extreme gradient boost, convolutional neural network, bidirectional long short-term memory, and convolutional neural network with multi-resolution blocks. The support vector machine achieved the best results with an F1-score of 0.81 (standard deviation: ±0.18), recall of 0.85±0.13, and precision of 0.79±0.22 in a leave-one-subject-out cross-validation. Our highly professional recordings and annotations provide a promising benchmark dataset for researchers to develop innovative machine learning approaches for precise HAR in free living.

Download Full-text

Data-driven predictions of a multiscale Lorenz 96 chaotic system using machine-learning methods: reservoir computing, artificial neural network, and long short-term memory network

Nonlinear Processes in Geophysics ◽

10.5194/npg-27-373-2020 ◽

2020 ◽

Vol 27 (3) ◽

pp. 373-389 ◽

Cited By ~ 7

Author(s):

Ashesh Chattopadhyay ◽

Pedram Hassanzadeh ◽

Devika Subramanian

Keyword(s):

Neural Network ◽

Machine Learning ◽

Short Term Memory ◽

Data Driven ◽

Reservoir Computing ◽

Short Term ◽

Term Memory ◽

Machine Learning Methods ◽

Long Short Term Memory ◽

Lorenz 96

Abstract. In this paper, the performance of three machine-learning methods for predicting short-term evolution and for reproducing the long-term statistics of a multiscale spatiotemporal Lorenz 96 system is examined. The methods are an echo state network (ESN, which is a type of reservoir computing; hereafter RC–ESN), a deep feed-forward artificial neural network (ANN), and a recurrent neural network (RNN) with long short-term memory (LSTM; hereafter RNN–LSTM). This Lorenz 96 system has three tiers of nonlinearly interacting variables representing slow/large-scale (X), intermediate (Y), and fast/small-scale (Z) processes. For training or testing, only X is available; Y and Z are never known or used. We show that RC–ESN substantially outperforms ANN and RNN–LSTM for short-term predictions, e.g., accurately forecasting the chaotic trajectories for hundreds of numerical solver's time steps equivalent to several Lyapunov timescales. The RNN–LSTM outperforms ANN, and both methods show some prediction skills too. Furthermore, even after losing the trajectory, data predicted by RC–ESN and RNN–LSTM have probability density functions (pdf's) that closely match the true pdf – even at the tails. The pdf of the data predicted using ANN, however, deviates from the true pdf. Implications, caveats, and applications to data-driven and data-assisted surrogate modeling of complex nonlinear dynamical systems, such as weather and climate, are discussed.

Download Full-text

Development and Evaluation of the Combined Machine Learning Models for the Prediction of Dam Inflow

Water ◽

10.3390/w12102927 ◽

2020 ◽

Vol 12 (10) ◽

pp. 2927

Author(s):

Jiyeong Hong ◽

Seoro Lee ◽

Joo Hyun Bae ◽

Jimin Lee ◽

Woon Ji Park ◽

...

Keyword(s):

Neural Network ◽

Machine Learning ◽

Random Forest ◽

Multilayer Perceptron ◽

Short Term Memory ◽

Learning Algorithms ◽

Absolute Error ◽

Machine Learning Algorithms ◽

Gradient Boosting ◽

Dam Inflow

Predicting dam inflow is necessary for effective water management. This study created machine learning algorithms to predict the amount of inflow into the Soyang River Dam in South Korea, using weather and dam inflow data for 40 years. A total of six algorithms were used, as follows: decision tree (DT), multilayer perceptron (MLP), random forest (RF), gradient boosting (GB), recurrent neural network–long short-term memory (RNN–LSTM), and convolutional neural network–LSTM (CNN–LSTM). Among these models, the multilayer perceptron model showed the best results in predicting dam inflow, with the Nash–Sutcliffe efficiency (NSE) value of 0.812, root mean squared errors (RMSE) of 77.218 m3/s, mean absolute error (MAE) of 29.034 m3/s, correlation coefficient (R) of 0.924, and determination coefficient (R2) of 0.817. However, when the amount of dam inflow is below 100 m3/s, the ensemble models (random forest and gradient boosting models) performed better than MLP for the prediction of dam inflow. Therefore, two combined machine learning (CombML) models (RF_MLP and GB_MLP) were developed for the prediction of the dam inflow using the ensemble methods (RF and GB) at precipitation below 16 mm, and the MLP at precipitation above 16 mm. The precipitation of 16 mm is the average daily precipitation at the inflow of 100 m3/s or more. The results show the accuracy verification results of NSE 0.857, RMSE 68.417 m3/s, MAE 18.063 m3/s, R 0.927, and R2 0.859 in RF_MLP, and NSE 0.829, RMSE 73.918 m3/s, MAE 18.093 m3/s, R 0.912, and R2 0.831 in GB_MLP, which infers that the combination of the models predicts the dam inflow the most accurately. CombML algorithms showed that it is possible to predict inflow through inflow learning, considering flow characteristics such as flow regimes, by combining several machine learning algorithms.

Download Full-text

A Machine Learning View on Momentum and Reversal Trading

Algorithms ◽

10.3390/a11110170 ◽

2018 ◽

Vol 11 (11) ◽

pp. 170 ◽

Cited By ~ 2

Author(s):

Zhixi Li ◽

Vincent Tam

Keyword(s):

Neural Network ◽

Machine Learning ◽

Stock Market ◽

Short Term Memory ◽

Predictive Ability ◽

Trading Strategies ◽

Machine Learning Techniques ◽

Support Vector ◽

Learning Approaches ◽

Learning Techniques

Momentum and reversal effects are important phenomena in stock markets. In academia, relevant studies have been conducted for years. Researchers have attempted to analyze these phenomena using statistical methods and to give some plausible explanations. However, those explanations are sometimes unconvincing. Furthermore, it is very difficult to transfer the findings of these studies to real-world investment trading strategies due to the lack of predictive ability. This paper represents the first attempt to adopt machine learning techniques for investigating the momentum and reversal effects occurring in any stock market. In the study, various machine learning techniques, including the Decision Tree (DT), Support Vector Machine (SVM), Multilayer Perceptron Neural Network (MLP), and Long Short-Term Memory Neural Network (LSTM) were explored and compared carefully. Several models built on these machine learning approaches were used to predict the momentum or reversal effect on the stock market of mainland China, thus allowing investors to build corresponding trading strategies. The experimental results demonstrated that these machine learning approaches, especially the SVM, are beneficial for capturing the relevant momentum and reversal effects, and possibly building profitable trading strategies. Moreover, we propose the corresponding trading strategies in terms of market states to acquire the best investment returns.

Download Full-text

Comparison of machine learning and deep learning techniques in promoter prediction across diverse species

PeerJ Computer Science ◽

10.7717/peerj-cs.365 ◽

2021 ◽

Vol 7 ◽

pp. e365

Author(s):

Nikita Bhandari ◽

Satyajeet Khare ◽

Rahee Walambe ◽

Ketan Kotecha

Keyword(s):

Neural Network ◽

Machine Learning ◽

Deep Learning ◽

Short Term Memory ◽

Regulatory Elements ◽

Gene Promoters ◽

Promoter Prediction ◽

Promoter Sequences ◽

Learning Techniques

Gene promoters are the key DNA regulatory elements positioned around the transcription start sites and are responsible for regulating gene transcription process. Various alignment-based, signal-based and content-based approaches are reported for the prediction of promoters. However, since all promoter sequences do not show explicit features, the prediction performance of these techniques is poor. Therefore, many machine learning and deep learning models have been proposed for promoter prediction. In this work, we studied methods for vector encoding and promoter classification using genome sequences of three distinct higher eukaryotes viz. yeast (Saccharomyces cerevisiae), A. thaliana (plant) and human (Homo sapiens). We compared one-hot vector encoding method with frequency-based tokenization (FBT) for data pre-processing on 1-D Convolutional Neural Network (CNN) model. We found that FBT gives a shorter input dimension reducing the training time without affecting the sensitivity and specificity of classification. We employed the deep learning techniques, mainly CNN and recurrent neural network with Long Short Term Memory (LSTM) and random forest (RF) classifier for promoter classification at k-mer sizes of 2, 4 and 8. We found CNN to be superior in classification of promoters from non-promoter sequences (binary classification) as well as species-specific classification of promoter sequences (multiclass classification). In summary, the contribution of this work lies in the use of synthetic shuffled negative dataset and frequency-based tokenization for pre-processing. This study provides a comprehensive and generic framework for classification tasks in genomic applications and can be extended to various classification problems.

Download Full-text

Analysis of comparative performance of deep learning models for sentiment analysis

The International FLAIRS Conference Proceedings ◽

10.32473/flairs.v34i1.128739 ◽

2021 ◽

Vol 34 (1) ◽

Author(s):

Mirza Murtaza

Keyword(s):

Neural Network ◽

Machine Learning ◽

Sentiment Analysis ◽

Language Processing ◽

Short Term Memory ◽

Neural Nets ◽

Test Accuracy ◽

Comparative Performance ◽

Text Filtering ◽

Percent Accuracy

Abstract Sentiment analysis of text can be performed using machine learning and natural language processing methods. However, there is no single tool or method that is effective in all cases. The objective of this research project is to determine the effectiveness of neural network-based architecture to perform sentiment analysis of customer comments and reviews, such as the ones on Amazon site. A typical sentiment analysis process involves text preparation (of acquired content), sentiment detection, sentiment classification and analysis of results. In this research, the objective is to a) identify the best approach for text preparation in a given application (text filtering approach to remove errors in data), and, most importantly, b) what is the best machine learning (feed forward neural nets, convolutional neural nets, Long Short-Term Memory networks) approach that provides best classification accuracy. In this research, a set of three thousand two hundred reviews of food related products were used to train and experiment with a neural network-based sentiment analysis system. The neural network implementation of six different models provided close to one-hundred percent accuracy of test data, and a decent test accuracy in mid-80%. The results of the research would be useful to businesses in evaluating customer preferences for products or services.

Download Full-text

Predicting Sentiment Polarity of Microblogs using an LSTM – CNN Deep Learning Model

International Journal of Engineering and Advanced Technology - Regular Issue ◽

10.35940/ijeat.f8933.088619 ◽

2019 ◽

Vol 8 (6) ◽

pp. 4368-4373

Keyword(s):

Neural Network ◽

Machine Learning ◽

Prediction Accuracy ◽

Short Term Memory ◽

Learning Model ◽

Supervised Machine Learning ◽

Machine Learning Model ◽

Proposed Model ◽

Network Approaches

In this paper we propose a novel supervised machine learning model to predict the polarity of sentiments expressed in microblogs. The proposed model has a stacked neural network structure consisting of Long Short Term Memory (LSTM) and Convolutional Neural Network (CNN) layers. In order to capture the long-term dependencies of sentiments in the text ordering of a microblog, the proposed model employs an LSTM layer. The encodings produced by the LSTM layer are then fed to a CNN layer, which generates localized patterns of higher accuracy. These patterns are capable of capturing both local and global long-term dependences in the text of the microblogs. It was observed that the proposed model performs better and gives improved prediction accuracy when compared to semantic, machine learning and deep neural network approaches such as SVM, CNN, LSTM, CNN-LSTM, etc. This paper utilizes the benchmark Stanford Large Movie Review dataset to show the significance of the new approach. The prediction accuracy of the proposed approach is comparable to other state-of-art approaches.

Download Full-text

Evaluation of Machine Learning-Driven Automatic Modulation Classifiers Under Various Signal Models

2020 Joint Rail Conference ◽

10.1115/jrc2020-8091 ◽

2020 ◽

Cited By ~ 1

Author(s):

Pejman Ghasemzadeh ◽

Subharthi Banerjee ◽

Michael Hempel ◽

Hamid Sharif ◽

Tarek Omar

Keyword(s):

Neural Network ◽

Machine Learning ◽

Communication Systems ◽

Short Term Memory ◽

Cognitive Radios ◽

Wireless Channel ◽

Support Vector ◽

Modulation Scheme ◽

Modulation Classification ◽

Non Linear

Abstract Automatic Modulation Classification (AMC) is becoming an essential component in receiver designs for next-generation communication systems, such as Cognitive Radios (CR). AMC enables receivers to classify an intercepted signal’s modulation scheme without any prior information about the signal. This is becoming increasingly vital due to the combination of congested frequency bands and geographically disparate frequency licensing for the railroad industry across North America. Thus, a radio technology is needed that allows train systems to adapt automatically and intelligently to changing locations and corresponding RF environment fluctuations. Three AMC approaches have been proposed in the scientific literature. The performance of these approaches depends especially on the particular environment where the classifiers are employed. In this work, the authors present a performance evaluation of the Feature-based AMC approach, as this is the most promising approach for railroads in real-time AMC operations under various different wireless channel environments. This study is done as the first one for railroads application where it considers different environments models including Non-Gaussian Class A noise, Multipath fast fading, and their combination. The evaluation is conducted for signals using a series of QAM modulation schemes. The authors selected the signal’s Cumulant statistical features for the feature extraction stage in this study, coupled with three different machine learning classifiers: Support Vector Machine (SVM), Deep Neural Network (DNN) and Recurrent Neural Network (RNN) utilizing long-short term memory (LSTM), in order to maintain control over the classifiers’ accuracy and computational complexity, especially for the non-linear cases. Our results indicate that when the signal model noise shows higher non-linear behavior, the RNN classifier on average achieves higher classification accuracy than the other classifiers.

Download Full-text

El Niño Index forecasting using machine learning techniques

10.5194/egusphere-egu21-5146 ◽

2021 ◽

Author(s):

Wanjiao Song ◽

Wenfang Lu ◽

Qing Dong

Keyword(s):

Neural Network ◽

Machine Learning ◽

Large Scale ◽

Short Term Memory ◽

Machine Learning Techniques ◽

Time Series Dataset ◽

The Pacific ◽

Forecast Efficiency ◽

Multiple Variables ◽

Lag Times

<p>El Ni&#241;o is a large-scale ocean-atmospheric coupling phenomenon in the Pacific. The interaction among marine and atmospheric variables over the tropical Pacific modulate the evolution of El Ni&#241;o. The latest research shows that machine learning and neural network (NN) have appeared as effective tools to achieve meaningful information from multiple marine and atmospheric parameters. In this paper, we aim to predict the El Ni&#241;o index more accurately and increase the forecast efficiency of El Ni&#241;o events. Here, we propose an approach combining a&#160;neural network technique with long short-term memory (LSTM) neural network to forecast El Ni&#241;o phenomenon. The attributes of model are resulted from physical explanation which are tested with the experiments and observations. The neural network represents the connection among multiple variables and machine learning creates models to identify the El Ni&#241;o events. The preliminary experimental results exhibit that training NN-LSTM model on network metrics time series dataset provides great potential for predicting El Ni&#241;o phenomenon at lag times of up to more than 6 months. &#160;</p>

Download Full-text

Comparação entre LSTM e CLCNN na detecção de requisições maliciosas em ataques na web

10.5753/sbseg.2021.17310 ◽

2021 ◽

Author(s):

Rafael Bosse Brinhosa ◽

Marcos A. Michels Schlickmann ◽

Eduardo da Silva ◽

Carlos Becker Westphall ◽

Carla Merkle Westphall

Keyword(s):

Neural Network ◽

Machine Learning ◽

Convolutional Neural Network ◽

Short Term Memory ◽

Short Term ◽

Term Memory ◽

Long Short Term Memory ◽

Cross Site

Com o uso de aplicações web em ambientes dinâmicos de computação em nuvem integrados com dispositivos IoT, os ataques de injeção de SQL e de XSS (Cross-Site Scripting) continuam causando problemas para a segurança. A detecção de requisições maliciosas a nível de aplicação representa um desafio na pesquisa, que está evoluindo usando técnicas de Machine Learning e redes neurais. Este trabalho apresenta a comparação entre duas arquiteturas de aprendizado de máquina usadas para detectar requisições web maliciosas: LSTM (Long Short-Term Memory) e CLCNN (Character-level Convolutional Neural Network). Os resultados demonstram que a CLCNN é a mais eficaz em todas as métricas, com uma acurácia de 98,13%, precisão de 99,84%, taxa de detecção em 95,66% e com um F1-score de 97,70%.

Download Full-text