Explainable Hybrid Intelligence Model for Accurate Anaemia Identification

Abstract

This project focuses on the development of an intelligent system for predicting anaemia using machine learning techniques. The system is implemented using Python in a Jupyter Notebook environment and utilizes a structured dataset containing patient health attributes. Various algorithms such as Decision Tree, K-Nearest Neighbors, Support Vector Machine, and Gradient Boosting are applied to analyze and classify anaemia conditions. A hybrid model is also proposed to improve prediction accuracy. The dataset is preprocessed through normalization and feature extraction techniques to enhance model performance. Evaluation metrics such as accuracy, precision, recall, and F1-score are used to compare models, where the hybrid model achieved the highest accuracy. Visualization techniques and interpretability tools like LIME and SHAP are used to explain predictions. Additionally, a Flask-based web application is developed to provide a userfriendly interface for real-time anaemia prediction.