AI-Driven Hematological Analysis for Proactive Dengue Diagnosis

Abstract

A method utilizing artificial intelligence for the early identification of dengue via complete blood count (CBC) data is added. Multiple function selection methodologies, which include Pearson Correlation, Recursive feature elimination (RFE) utilizing Random forest, SelectKBest, Chi-square (Chi2), and ExtraTree, are hired to ascertain the most pertinent features. various machine learning and deep learning algorithms are utilized, including Logistic Regression, support Vector machine (SVM), Naive Bayes, Random forest, AdaBoost, XGBoost, Multi-Layer Perceptron (MLP), LightGBM, as well as ensemble methods such as a Stacking Classifier (XGB + LR + MLP with LightGBM) and voting Classifier (Boosted decision Tree + ExtraTree). Deep learning architectures, such as artificial Neural Networks (ANN), Convolutional Neural Networks (CNN), Gated Recurrent units (GRU), Bidirectional long short-term memory (Bi-LSTM), Feedforward Neural Networks (FNN), Transformers, and hybrid fashions like CNN + LSTM, are hired. Integrating predictions from distinct models via ensemble approaches improves robustness and precision, with the voting Classifier attaining ninety eight% accuracy and F1 rating. The implementation of hybrid models, in particular CNN mixed with LSTM, enhances the system's performance. The methodology is based for user engagement and verification the use of a Flask-based totally interface with authentication, guaranteeing accessibility and security while preserving excessive predictive accuracy.