Raisin Grain Classification Using Machine Learning Models

This study presents machine learning models developed to classify two different species of raisins grown in Turkey. This study uses the Raisin Dataset from the University of California Irvine Machine Learning Repository. The Raisin Dataset contains 7 morphological features of 900 raisin grains. This study presents an efficient way of pre-processing the raisin data using Feature Selection with Random Forest to determine the most important features. By only using one feature that was found to be the most important, the best model (Multi-Layer Perceptron) yields an accuracy of 94.07%, outperforming previous models used to classify raisin grains with this dataset.