EZE, IZUCHUKWU VALENTINE (2026) DESIGN OF A WEB-BASED MACHINE LEARNING SYSTEM FOR REAL-TIME CROP YIELD PREDICTION IN DROUGHT-PRONE NIGERIA. Other thesis, GODFREY OKOYE UNIVERSITY, ENUGU,.
![[img]](https://eprints.gouni.edu.ng/style/images/fileicons/text.png) |
Text
VALENTINE PROJECT COMPLETE.pdf Download (2MB) |
Abstract
The traditional approaches to crop yield estimation fails in capturing the nonlinear relationship between the environmental parameters. Therefore, a machine learning based approach has been employed in crop yield prediction based on the significant environmental factors like weather and soil parameters. The methodology involves the use of supervised machine learning techniques particularly Decision Trees. It is implemented in form of a web-based application that is interactive and real-time in nature. After the collection of the data regarding ecology, certain preprocessing steps have been taken including normalization and feature selection. The collected data has been divided into two parts namely training dataset and testing dataset. A number of machine learning models were implemented and evaluated including Linear Regression, Random Forest, XGBoost, K-Nearest Neighbors and Decision Tree. The results of performances indicated that Linear Regression was the best performing method among all of the applied models with maximum accuracy of 91.3% and minimum Mean Squared Error of 0.25. XGBoost and Random Forest were the secondbest methods that were very close to each other. The Decision Tree Method was comparatively less accurate with 81.5% but was competitive too in performance making it a very useful tool for yield estimation.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Computing And Information Technology (FACIT) |
| Depositing User: | Nnenna Ayo |
| Date Deposited: | 24 Jul 2026 15:21 |
| Last Modified: | 24 Jul 2026 15:21 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6017 |
Actions (login required)
 |
View Item |
