Article: HML-VBSP: a hybrid machine learning framework for predicting multiple vector-borne diseases simulation using soft voting strategy and hyper parameter tuning systems Journal: International Journal of Engineering Systems Modelling and Simulation (IJESMS) 2026 Vol.17 No.4 pp.195 - 209 Abstract: Dengue, yellow fever, chikungunya, and Zika are vector-borne illnesses that are becoming more common as the world's population grows. Environmental variables, such as temperature and precipitation, are frequently incorporated into infectious disease models. Early warning of disease outbreaks may be provided by combining forecasting models with increasing computer capability and better AI technologies. Then, using a voting classifier with a 'hard' voting strategy, the proposed HML-VBDP model for vector-borne disease prediction is constructed by combining a random forest (RF), a support vector classifier (SVC), and a gradient boosting (GB) classifier. To optimise parameters such as the learning rate, number of estimators for GB, regularisation parameter (C), kernel coefficient (gamma) for SVC, and maximum depth and number of estimators for RF, we use RandomizedSearchCV to tune each classifier's hyperparameters before training the model. To train the HML-VBDP model, we utilise the training data. Then, to check its performance, we use the testing data. The efficacy of the model is assessed using evaluation measures including recall, accuracy, precision, F1-score, ROC curve, and confusion matrix. Inderscience Publishers - linking academia, business and industry through research

Title: HML-VBSP: a hybrid machine learning framework for predicting multiple vector-borne diseases simulation using soft voting strategy and hyper parameter tuning systems

Authors: K. Kavitha; T. Prabhu

Addresses: Department of Computer Applications, Dr. M.G.R. Educational and Research Institute, Chennai, Tamil Nadu, India ' Department of Computer Applications, Dr. M.G.R. Educational and Research Institute, Chennai, Tamil Nadu, India

Abstract: Dengue, yellow fever, chikungunya, and Zika are vector-borne illnesses that are becoming more common as the world's population grows. Environmental variables, such as temperature and precipitation, are frequently incorporated into infectious disease models. Early warning of disease outbreaks may be provided by combining forecasting models with increasing computer capability and better AI technologies. Then, using a voting classifier with a 'hard' voting strategy, the proposed HML-VBDP model for vector-borne disease prediction is constructed by combining a random forest (RF), a support vector classifier (SVC), and a gradient boosting (GB) classifier. To optimise parameters such as the learning rate, number of estimators for GB, regularisation parameter (C), kernel coefficient (gamma) for SVC, and maximum depth and number of estimators for RF, we use RandomizedSearchCV to tune each classifier's hyperparameters before training the model. To train the HML-VBDP model, we utilise the training data. Then, to check its performance, we use the testing data. The efficacy of the model is assessed using evaluation measures including recall, accuracy, precision, F1-score, ROC curve, and confusion matrix.

Keywords: hyper parameter tuning systems; vector-borne disease; VBD; random forest; support vector machine; SVM; gradient boosting; vector-borne diseases simulation.

DOI: 10.1504/IJESMS.2026.155147

International Journal of Engineering Systems Modelling and Simulation, 2026 Vol.17 No.4, pp.195 - 209

Received: 06 Jun 2024
Accepted: 19 Sep 2025
Published online: 28 Jul 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: HML-VBSP: a hybrid machine learning framework for predicting multiple vector-borne diseases simulation using soft voting strategy and hyper parameter tuning systems

Keep up-to-date