Comparative Analysis of Cutaneous Leishmaniasis Future Forecasting Using Supervised Machine Learning Models

EasyChair Preprint no. 13401

Abstract

Cutaneous leishmaniasis (CL) represents a considerable public health problem, with its incidence influenced by a complex interplay of ecological and socio-environmental variables. Forecasting its incidence accurately is pivotal for the strategizing of control measures and optimal resource distribution. This study aims to predict the incidence of CL through the application of supervised machine learning techniques to historical data spanning from 2005 to 2022. Three models were employed including, AutoRegressive Integrated Moving Average (ARIMA), Linear Regression (LR), and Support Vector Machine (SVM), and their forecasting performance was assessed using a suite of statistical metrics. The SVM model outperformed the others, demonstrating the lowest error rates and strongest predictive performance, particularly adept at navigating the non-linear epidemiological patterns of CL. The ARIMA model offered balanced results, whereas the LR model, although simplest, was less precise. The SVM model was then applied to predict CL incidence rates over the next 18 years in six countries known to have historically high incidence rates, incorporating climate data into their analysis. Our research highlights the efficacy of machine learning in epidemiological predictions and suggests that SVM models hold substantial promise for future public health applications, providing a robust approach for the forecasting of CL incidences. These insights are crucial for public health authorities to proactively manage and prevent CL outbreaks, indicating a step forward in the application of advanced analytics in disease surveillance and response planning.

Keyphrases: ARIMA, Cutaneous Leishmaniasis, Forecasting, LR, machine learning, SVM

@Booklet{EasyChair:13401,
  author = {Hasnaa Talimi and Imane El Idrissi Saik and Meryem Lemrani and Rachida Fissoune},
  title = {Comparative Analysis of Cutaneous Leishmaniasis Future Forecasting Using Supervised Machine Learning Models},
  howpublished = {EasyChair Preprint no. 13401},

  year = {EasyChair, 2024}}