Finding Recommendation from Crop Dataset using Map Reduce Machine Learning Algorithm

Finding Recommendation from Crop Dataset using Map Reduce Machine Learning Algorithm

Abstract — The country's economic prosperity and expansion are mostly attributed to agriculture. Approximately 17% of India's GDP comes from agriculture and related industries, and 70% of Indians still choose this as their primary line of work. The farmers' failure to select the proper crop for cultivation is the primary and most significant hindrance to crop productivity. Farmers can benefit greatly from this kind of technology because it is portable, simple to use, and doesn't require a lot of hardware. It becomes much more crucial in India, since the agricultural research and development agency lacks sufficient facilities. Random Forest and Naive Bayes are the independent base learners employed in the ensemble model. The crop recommendation algorithm classifies the input by separating the suggested crop types, Kharif and Rabi, from the soil dataset.. The dataset includes samples of the average rainfall and surface temperature as well as the physical and chemical properties particular to the soil. Combining the independent base learners yields an average classification accuracy of 99.91 percent.

Keywords – Map Reduce, Naïve Bayes, Random Forest, Decision Tree , crop recommendation