© 2020, IJCSE All Rights Reserved 57 International Journal of Computer Sciences and Engineering Open Access Research Paper Vol.8, Issue.6, June 2020 E-ISSN: 2347-2693 Forecasting novel COVID-19 confirmed cases in India using Machine Learning Methods Saroj S. Date 1* , Sachin N. Deshmukh 2 1 Dept. of Computer Science & Engineering, Jawaharlal Nehru Engineering College, MGM University, Aurangabad (MS), India 2 Dept. of Computer Science & Information Technology, Dr. Babasaheb Ambedkar Marathwada University, Aurangabad (MS) India *Corresponding Author: sarojdate@jnec.ac.in DOI: https://doi.org/10.26438/ijcse/v8i6.5762 | Available online at: www.ijcseonline.org Received: 26/May/2020, Accepted: 20/June/2020, Published: 30/June/2020 Abstract— Nowadays, there is a very adverse impact on economic, cultural, social and almost all fields in the world because of Covid-19. The Covid-19 term is described as -'CO' for corona, 'VI' for virus, and 'D' for disease. It is an infectious disease caused by severe acute respiratory syndrome which is transmitted through respiratory droplets and contact routes. Since December 2019, corona-virus disease (COVID-19) has out-broke from the country China. Till now, more than 78, 23, 289 people are infected and more than 4 Lakhs of deaths have been caused worldwide. Unfortunately, the number of infections and deaths are still increasing rapidly which has put the world in a different state. Artificial Intelligence can play a key role to infection forecasting in national and provincial levels in many countries. The objective of this study is to use machine learning methods to forecast the number of cases for the next 2 weeks, i.e. till 30 th June 2020. The data was collected from 22 nd January to 15 th June 2020 by nationally recognized sources. The data file contains the cumulative count of confirmed, death and recovered cases of COVID-19 from different countries from the date 22 nd January 2020.In this study, the outbreak of this disease has been analyzed for India till 15 th June 2020 and predictions have been made for the number of cases for the next two weeks. Keywords—Covid-19, Corona, Corona Virus, Machine Learning, Forecasting, Artificial Intelligence, time series forecasting I. INTRODUCTION On 31 st December 2019, the novel Corona virus, known as COVID-19 was reported in Wuhan, China for the very first time. Corona viruses are the infectious virus which has adverse affect on the respiratory system of humans. The symptoms of COVID-19 may or may not be visual in infected individual, therefore the spread rate can be faster. Till now, effective and well-tested vaccine against CoVID- 19 has not been invented, only precautions are the safety measures. Though the continuous efforts are going on , the virus has managed to spread in most of the territories in the world and World Health Organization (WHO) has announced COVID-19 as Pandemic. Most of the countries in the world are working cooperatively and openly to bring this situation under control. Data scientists and data mining researchers can play an important role during these types of situations. They can integrate the related data and technology to better understand the virus and its characteristics, which can help in taking right decisions and concrete plan of actions. As per the daily situation report of WHO, as on 15 th June 2020 the COVID-19 transmission scenario reports 78,23,289 confirmed cases with 4,31,541 deaths globally. Data mining is a technology, developing with database as well as artificial intelligence. It is a processing procedure of extracting credible and effective novel techniques and understandable patterns from the database [1]. Artificial intelligence (AI) is a field of programming building which gives PCs an ability to learn without being unequivocally modified [2]. AI models can be used for estimating and predicting spread rate, so AI is one of the beneficial tools to fight against pandemic like COVID-19. The forecasting analysis is done by using the algorithms like ANN and time-series [3]. In this paper we are using, time-series algorithm. According to K. Krishna Rani Samal et al., approaches like SARIMA and Prophet can be used for forecasting based on historical data. They concluded that both the SARIMA and prophet model provides a good quality of accuracy. However, the best approach is the prophet model on log transformation which has the least minimum RMSE, MSE value [4]. This model is developed by Facebook, available in python and R. The main contribution of this research paper is forecasting of COVID-19 for the next two weeks i.e. till 30 th June 2020 using Prophet Model. In this study, the data was analyzed from 22 nd January 2020 to 15 th June 2020.