Detecting Sugarcane Crop Yield using Decision Tree Classifier in the District of Muzaffarnagar

Abstract

The district of Muzaffarnagar is the highest sugarcane producing district in Uttar Pradesh and therefore is an important industrial district as well. The district is part of Western UP and it shares the problems of the sugar industry elsewhere in the state: unpredictable demands and crop failures. In this context, predicting sugarcane demand and informing its production can turn to be just the key to solve some of the problems the industry faces. The existing crop forecasting method for the cultivation of sugarcane used in UP relies, to a large degree, on subjective details, centred on the expertise of engineers in the sugar and alcohol field and on information on input demand in the supply chain. The measurement of the utility of the sample detection using NDVI images from the SPOT sensor used in the sensor's determination over the ECMWF model was possible to infer the official productivity data reported in the previously selected municipalities and harvest. Significant features of the municipal productivity of a given village is listed in a decision tree, and out of the combinations of attributes the corresponding municipal productivity is rated as "Normal" on the average urban productivity scale. Using data from the NDVI time-series between 2013 to 2020, we can discern the three classes of productivity in the meanwhile. Findings indicate that productivity in January ranked as less than mean, mean, and more than mean. The findings were more successful for the class Vegetation, the participants of which were permitted to conclude about the pattern of the average federal productivity prior to.