FARMER: Forecasting Agricultural Results with Machine Learning Evaluation and Regression

Agriculture plays an important role, in Indias economy providing support to a portion of the population and contributing significantly to the GDP. It is crucial to predict crop yields in order to make decisions and promote sustainable agricultural practices. To achieve this a research study proposes an approach that utilizes machine learning techniques such as Decision Tree, Support Vector Machine (SVM) Random Forest, Multiple Linear Regression and Partial Least Squares Regression. The study carefully evaluates the performance of these models. Reveals some findings while comparing them to determine the best model. Multiple Linear Regression achieves the accuracy at 81.2% followed by Decision Tree at 83.5% SVM at 92.1% Random Forest at 91.5%. Partial Least Squares Regression at 85.3%. Notably SVM emerges as the performing model demonstrating its potential in improving precision, for predicting agricultural yields. Ultimately an Android application is developed using Kotlin and Android Studio for an user experience.