Abstract Agriculture isone of the main income producing sectors in India. A mixture of biological,seasonal and economic factors manipulate the production of the crops, but randomchanges causes a great loss to the farmers. By applying appropriatemathematical or statistical methodologies in the soil and weather data, theserisks can be computed. Machine learning helps to forecast the crop yield byobtaining the valuable information from the agricultural data to make a decision on the crop for farmers, sothat they can plant for the future which leads to huge profit. In this paper,we present a detailed study on the various machine learning algorithms thatfacilitate to forecast the crop yield.
Keywords:Machine Learning, Data Mining, Crop yield, Forecasting, Agriculture I. INTRODUCTION Agricultureforms the important source for food security. It assists human beings to growthe best food crops. Rice and wheat is the primary food in India. Indian farmers grows the followingfoods such as rubber, cotton, potatoes, pulses sugarcane, oilseeds.
Agricultureis depended by 70 per cent of the rural family. Total GDP of 17% is contributedin agriculture. 60% employment isprovided over the population. A machine learning technique helps to make decisionsautomatically by detecting pattern from the past data and generalizing it onthe future data.
Machine learningalgorithms are anticipated to replace 25% of the occupation across the world inthe next 10 years. The following are the 3major categories of Machine Learning algorithmsIn Supervised Learning, we have inputand output variables and the algorithm create a function that calculates theoutput based on given input variables. Regressionand Classification are the two parts: Some examples includeLinear Regression, Decision Trees, Random Forest, k nearest neighbours, SVM,Gradient Boosting Machines (GBM), Neural Network etc. In Unsupervised learning, only input data is presentand there is no corresponding output variable.
It can also be classified intotwo groups, namely Cluster analysis and Association. Some examples would bek-means clustering, hierarchical clustering, PCA, Apriori algorithm, etc. In Reinforcementlearning, the machine is given training to make accurate decisions from theseactions and tries to capture the best possible knowledge.
Some examples are Weather forecast, Speech Recognition, Game playing, face detection/Facerecognition, Genetics and agriculture.The rest of the paper is organized asfollows: Chapter II explains the methods of Machine Learning. Chapter IIIdescribes about the applications of Machine Learning used in agriculturedomain. Chapter IV analyses the outcomes. Chapter V discusses the conclusion.