Comparative Analysis of ML Algorithms for Drought for Prediction

Many droughts can be predicted up to a month in advance, and in rare cases it may be possible to predict drought conditions more than a year in advance. However, the complexity of Earth’s climate makes drought forecasting very difficult. To overcome these severe problems, the urgent need is to monitor or predict the drought severity of lands. Some droughts are easier to predict than others, and some still take us by surprise, but drought forecasting continues to improve as more data are collected and better models are produced. With continuous increase in the drought, most of the human life will get into panic situation. Now a days Artificial Intelligence is increasing its roots in every domain. So by using the subset machine learning from the Artificial Intelligence we are going to forecast the drought. Traditional time series analysis methods such as vector auto regression (Vector Auto Regression), arima (Auto Regressive Integrated Moving Averages) are implemented in this project. And also we implemented the most powerful and fully automated algorithm prophet which was developed by Facebook’s Data science team.

 KEYWORDS: VAR, ARIMA, prophet, drought..,