Forecasting COVID-19 Pandemic using Prophet, LSTM, hybrid GRU-LSTM, CNN-LSTM, Bi-LSTM and Stacked-LSTM for India

The COVID-19 Pandemic has been around for four years and remains a health concern for everyone. Although things are somewhat returning to normal, increased incidence of COVID-19 cases in some regions of the world (such as China, Japan, France, South Korea, etc.) has bred worry and anxiety in world, including India. The scientific community, which includes governmental organizations and healthcare facilities, was eager to learn how the COVID-19 Pandemic would develop. The current work makes an attempt to address this question by employing cutting-edge machine learning and Deep Learning algorithms to anticipate the daily incidence of COVID-19 for India over the course of the next six months. For the purpose famous timeseries algorithms were implemented including LSTM, Bi-Directional LSTM and Stacked LSTM and Prophet. Owing to success of hybrid algorithms in specific problem domains- the present study also focuses on such algorithms like GRU-LSTM, CNN-LSTM and LSTM with Attention. All these models have been trained on timeseries dataset of COVID-19 for India and performance metrics are recorded. Of all the models, the simplistic algorithms have performed better than complex and hybrid ones. Owing to this best result was obtained with Prophet, Bidirectional LSTM and Vanilla LSTM. The forecast reveals flat nature of COVID-19 case load for India in future six months. . © 2023 IEEE.

A comparative study for COVID-19 cases forecasting with loss function as AIC and MSE in RNN family and ARIMA

Forecasting COVID-19 incidents is a trending research study in today's world. Since Machine learning models have been occupied in forecasting recently, this study focus on comparing statical and machine learning models such as ARIMA, RNN, LSTM, Seq2Seq, and Stacked LSTM. The performances were evaluated using two loss functions, namely, AIC and RMSE. The results showed that RNN performs with the lowest RMSE with-49.5% compared with the ARIMA. Seq2Seq scored the highest correlation of determination (R2) with 0.92. © 2022 IEEE.

An Investigation into Crime Forecast Using Auto ARIMA and Stacked LSTM

The severity of criminal activities which cause both physical and psychological damage has been increasing at an alarming rate across the globe. Realizing the significance of this problem, law enforcement agencies have developed several strategies to prevent crimes. Being slow-paced and ineffective in most cases, these prevention strategies are not robust enough to contribute in predicting crime trends for an early prevention. In this paper, we propose a regression-based model that incorporates temporal, statistical relationships and other relevant information about the data to forecast crime trends. Since, seasonal information is a powerful inclusion in an application of time series pattern, we use two popular regression methods, including an extended Autoregressive Integrated Moving Average (Auto ARIMA) and stacked Long Short-Term Memory (LSTM) to analyze crime patterns, specifically during the Covid-19 pandemic lockdown, and generate forecasts. We experimented our methods on London Crime Dataset and obtained some interesting results which can not only be useful to take necessary precautions, but also analyze crime patterns during the period of pandemic lockdowns for generating useful guidelines regarding citizens' life styles and hence, contribute to reducing the crime rates accordingly. © 2022 IEEE.

Forecasting of COVID-19 cases using deep learning models: Is it reliable and practically significant?

The ongoing outbreak of the COVID-19 pandemic prevails as an ultimatum to the global economic growth and henceforth, all of society since neither a curing drug nor a preventing vaccine is discovered. The spread of COVID-19 is increasing day by day, imposing human lives and economy at risk. Due to the increased enormity of the number of COVID-19 cases, the role of Artificial Intelligence (AI) is imperative in the current scenario. AI would be a powerful tool to fight against this pandemic outbreak by predicting the number of cases in advance. Deep learning-based time series techniques are considered to predict world-wide COVID-19 cases in advance for short-term and medium-term dependencies with adaptive learning. Initially, the data pre-processing and feature extraction is made with the real world COVID-19 dataset. Subsequently, the prediction of cumulative confirmed, death and recovered global cases are modelled with Auto-Regressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Stacked Long Short-Term Memory (SLSTM) and Prophet approaches. For long-term forecasting of COVID-19 cases, multivariate LSTM models is employed. The performance metrics are computed for all the models and the prediction results are subjected to comparative analysis to identify the most reliable model. From the results, it is evident that the Stacked LSTM algorithm yields higher accuracy with an error of less than 2% as compared to the other considered algorithms for the studied performance metrics. Country-specific analysis and city-specific analysis of COVID-19 cases for India and Chennai, respectively, are predicted and analyzed in detail. Also, statistical hypothesis analysis and correlation analysis are done on the COVID-19 datasets by including the features like temperature, rainfall, population, total infected cases, area and population density during the months of May, June, July and August to find out the best suitable model. Further, practical significance of predicting COVID-19 cases is elucidated in terms of assessing pandemic characteristics, scenario planning, optimization of models and supporting Sustainable Development Goals (SDGs).