Evaluation of Machine Learning Algorithms for Crop Yield Prediction
Renuka1, Sujata Terdal2

1Renuka, Computer Science and Engineering, PDACE, Kalaburagi, India.
2Dr. Sujata Terdal, Computer Science and Engineering, PDACE, Kalaburagi, India.
Manuscript received on July 01, 2019. | Revised Manuscript received on July 22, 2019. | Manuscript published on August 30, 2019. | PP: 4082-4086 | Volume-8 Issue-6, August 2019. | Retrieval Number: F8640088619/2019©BEIESP | DOI: 10.35940/ijeat.F8640.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Agriculture plays a significant role in the growth of the national economy. It relay on weather and other environmental aspects. Some of the factors on which agriculture is dependent are Soil, climate, flooding, fertilizers, temperature, precipitation, crops, insecticides and herb. The crop yield is dependent on these factors and hence difficult to predict. To know the status of crop production, in this work we perform descriptive study on agricultural data using various machine learning techniques. Crop yield estimates include estimating crop yields from available historical data such as precipitation data, soil data, and historic crop yields. This prediction will help farmers to predict crop yield before farming. Here we are utilizing three datasets like as clay dataset, precipitation dataset, and production dataset of Karnataka state, then we structure an assembled data sets and on this dataset we employ three different algorithms to get the genuine assessed yield and the precision of three different methods. K-Nearest Neighbor(KNN), Support Vector Machine(SVM), and Decision tree algorithms are applied on the training dataset and are tested with the test dataset, and the implementation of these algorithms is done using python programming and spyder tool. The performance comparison of algorithms is shown using mean absolute error, cross validation and accuracy and it is found that Decision tree is giving accuracy of 99% with very less mean square error(MSE). The proposed model can exhibit the precise expense of assessed crop yield and it is mark like as LOW, MID, and HIGH. 
Keywords: Descriptive analytics, Agriculture, Machine learning, K-Nearest Neighbour, Support Vector Machine, Decision tree.