Machine learning has evidenced to be a significant field of study over the past decade by solving several complicated real-world issues. Machine Learning applications have long been utilized in many application domains that require distinguishing and prioritizing negative factors in danger. This paper demonstrates the ability of the ML model to predict the COVID-19 epidemic, which is currently considered a potential threat to mankind. Several Machine learning techniques are being popularly used to handle forecasting problems. Most of the models predict the newly infected cases only but the death rate forecast and recovery rate forecast are not done. The forecasts are not applied to the particular period of time which is difficult to understand. Evaluation of multiple ML models not done, hence it is difficult to find out the best model. In this study classification and regression techniques are applied on four ML models such as Linear Regression, Least Absolute Shrinkage Selection Operator ,Exponential Smoothing ,Support Vector Machine to forecast number of confirmed cases, deaths, recoveries of the covid-19 pandemic for the time period of 10 days. Evaluation of models has been done to find the best model.