Animal Health Prediction Using Hybrid KNN Based Vector Neighbor Classification Model: A Machine Learning Approach

Abstract

Animal condition disease prediction is a crucial task in veterinary medicine, and accurate prediction can significantly improve animal health and reduce economic losses. This paper proposed a novel hybrid classification model that combines an Improved Fuzzy-based feature selection process with a Hybrid KNN-based Vector Neighbor Classification model to enhance animal condition disease prediction. The proposed model aims to improve the analysis of both normal and diseased categories by leveraging the strengths of fuzzy logic and machine learning. The Improved Fuzzy feature selection process utilizes a fuzzy-based algorithm to select relevant features from a high-dimensional dataset, reducing dimensionality and improving model performance. The Hybrid KNN-based Vector Neighbor Classification model integrates the benefits of K-Nearest Neighbors (KNN) and Vector Neighbor Classification (VNC) to classify animal conditions into normal and diseased categories. The proposed hybrid model is evaluated on a dataset of animal health records, and the results demonstrate improved accuracy and robustness compared to existing classification models. The experimental results of the proposed HKNN VNC method are evaluated on a real-world Animal condition classification dataset, achieving impressive performance metrics: Accuracy: 98.62%, Precision: 98.62%, Recall: 100%, and F1-score: 99.30%. These results demonstrate the effectiveness of the HKNN-VNC approach in accurately classifying animal conditions.