Analysis on prediction of rubber crop production using machine learning

Pang, Hui Jing (2021) Analysis on prediction of rubber crop production using machine learning. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Preview

Text (Full Text) Analysis on prediction of rubber crop production using machine learning.pdf - Submitted Version Download (5MB) | Preview

Abstract

Agriculture sectors are important for the economic growth of countries because it provides job opportunities for majority of population from developing countries. Agriculture provides a significant amount of raw materials to industries like sugar, cotton, palm oil, natural rubber, and so on. Agriculture is the backbone of Malaysia’s economy and to date, Malaysia has exported rubber products to various countries. Thus, prediction system of rubber crop production is crucial for making financial decisions earlier when crop production shortage is estimated by system. However, the human based prediction methods are involved time and labour consuming. Crop cut method may conduct measurement errors during weighing process and prone to overestimate the production of crop. In order to solve this problem, a prediction system using machine learning is introduced. This study is implemented and proved the ability to predict the rubber crop production of several states in Melaka, Perak, Pahang and Johor. A various types of machine learning algorithms are applied such as Random Forest, Decision Tree, Linear Regression and Neural Network. The Mean Square Error (MSE) and Mean Absolute Error (MAE) are used as an indicator to evaluate the performance of prediction models. At the end of this study, Linear Regression is able to provide accurate prediction results with the lowest MAE value in comparison with the abovementioned the machine learning algorithms.

Actions (login required)

View Item