To address the problem of predicting crop yields across Karnataka, a machine learning-based system is developed. The model will use factors such as Season, Crop Type, Area, Fertilizer, Pesticide, Production, and Rainfall to forecast yield accurately. The dataset will be preprocessed by cleaning, encoding, and normalizing features. Important parameters are selected through correlation analysis to enhance model efficiency. Four algorithms — Linear Regression, Decision Tree, Random Forest, and Support Vector Regression (SVR) — will be trained and compared. This solution provides an effective tool for farmers and policymakers to make informed decisions, optimize resources, and improve agricultural productivity across Karnataka.