Automated farming prediction

Farming is a predominantly manual process. The incorporation of any form of automation through the means of machine learning algorithms is still in incipient stage. This paper aims to introduce a fundamental approach to inaugurate the use of machine learning systems in the farming process. A comparative study between machine learning algorithms had been carried out in order to determine which algorithm is the most accurate in predicting the best crop for a particular land. Here, the best crop signifies the crop which had the most increase in terms of yield per unit area compared to previous years. This will ensure proper crop allocation throughout the country since optimal production for each crop will be secured. It will also increase the farmer's revenue margin. The study focuses on six major crops of Bangladesh which are Aman rice, Aus rice, Boro rice, Potato, Wheat and Jute. The algorithms that were used are Multiple Linear Regression (MLR) and K-Nearest Neighbor Regression (KNNR). Multiple Linear Regression (MLR) gave the most accurate results during the analysis and was incorporated into an android application. The android application system is also able to prepare a schedule of the complete farming process for a chosen crop, e.g. the correct time to apply fertilizers, irrigation, etc.