Mulberry Leaf Yield Prediction Using Machine Learning Techniques

TL;DR

This paper compares machine learning models to predict Mulberry leaf yield based on soil parameters, highlighting the effectiveness of Random Forest Regression for agricultural planning.

Contribution

It introduces a comparative analysis of ML techniques for Mulberry yield prediction, emphasizing the superior performance of Random Forest Regression.

Findings

Random Forest Regression outperforms other models

Soil parameters significantly influence yield predictions

ML models can assist farmers in planning agricultural activities

Abstract

Soil nutrients are essential for the growth of healthy crops. India produces a humungous quantity of Mulberry leaves which in turn produces the raw silk. Since the climatic conditions in India is favourable, Mulberry is grown throughout the year. Majority of the farmers hardly pay attention to the nature of soil and abiotic factors due to which leaves become malnutritious and thus when they are consumed by the silkworm, desired quality end-product, raw silk, will not be produced. It is beneficial for the farmers to know the amount of yield that their land can produce so that they can plan in advance. In this paper, different Machine Learning techniques are used in predicting the yield of the Mulberry crops based on the soil parameters. Three advanced machine-learning models are selected and compared, namely, Multiple linear regression, Ridge regression and Random Forest Regression (RF). The experimental results show that Random Forest Regression outperforms other algorithms.