Prediction of COVID-19 confirmed, death, and cured cases in India using random forest model

Abstract

A novel coronavirus (SARS-CoV-2) is an unusual viral pneumonia in patients, first found in late December 2019, latter it declared a pandemic by World Health Organizations because of its fatal effects on public health. In this present, cases of COVID-19 pandemic are exponentially increasing day by day in the whole world. Here, we are detecting the COVID-19 cases, i.e., confirmed, death, and cured cases in India only. We are performing this analysis based on the cases occurring in different states of India in chronological dates. Our dataset contains multiple classes so we are performing multi-class classification. On this dataset, first, we performed data cleansing and feature selection, then performed forecasting of all classes using random forest, linear model, support vector machine, decision tree, and neural network, where random forest model outperformed the others, therefore, the random forest is used for prediction and analysis of all the results. The K-fold cross-validation is performed to measure the consistency of the model.

Learning From Biological and Computational Machines: Importance of SARS-CoV-2 Genomic Surveillance, Mutations and Risk Stratification

“…Upon close observation of the characteristics of the other models, it became clear that a simpler model such as logistic regression (F1 score = 0.51) cannot capture the complexity of the dataset, even after selecting the best features and tuning the hyperparameters. In various other studies with similar models ( Prakash, 2020 ; Gupta V. K. et al., 2021 ), it was observed that logistic regression underperformed. In our study, random forest, multilayer perceptron, XGBoost, and SVM performed much better and had similar scores, so we can use the results of the four combined to arrive at a prediction.…”

“…To avoid information leaking into the test set and overfitting of data, nested CV effectively uses a series of different train–test set splits in each iteration. Nested CV is the preferred way to evaluate and compare tuned ML models and has been used before in clinical settings ( Gupta V. K. et al., 2021 ). Figure 2 demonstrates the workflow of the nested CV algorithm.…”

The role of European health system characteristics in affecting Covid 19 lethality during the early days of the pandemic

“…The main limitation of the study is that we analyze, due to lacking data, only 16 European countries. However, unlike previous studies are limited in size examining just a single country 36 – 38 , 40 , our goal is to use the data from 16 European countries to analyze with ML techniques the global effect of different health system characteristics on COVID-19 lethality during the early days of the pandemic.…”

Evaluating Temporal Bias in Time Series Event Detection Methods

“…Besides, ELM tends to present fewer optimization constraints, greater scalability, superior results for generalization performance, and lower learning speed concerning the others models. Gupta et al [2021] analyzed different cases of COVID-19 in India (confirmed, killed, and recovery), using machine learning models: Random Forest, Linear Model, SVM, Decision Tree e Neural Network. The results revealed that the Random Forest presented a superior performance compared to the other evaluated models.…”