Dengue outbreak prediction in Sri Lanka using machine learning techniques



Table of content

  1. Abstract
  2. Related works
  3. Experiment Setup and Implementation
  4. Conclusion
  5. Publications
  6. Links


This research focuses on the prediction of dengue fever outbreaks, which have become a significant public health concern in many tropical and subtropical regions worldwide. The study explores the complex and dynamic nature of the disease, considering various factors that influence its spread. By utilizing vector populations, urbanization, climate parameters, and other relevant factors, the research aims to develop effective strategies for early detection and prevention of dengue outbreaks in Sri Lanka with the aid of machine learning techniques

image info

This image depicts the dengue trend in sri lanka during 1989 to 2021
image info

Here are a few related works we discovered after reviewing various research pages.

Epidemiological Prediction Using Machine Learning

Machine learning models are increasingly utilized for epidemiological prediction, leveraging historical data to classify future dengue incidence as either high or low. Techniques like Fuzzy Association Rule Mining have been pivotal in extracting intricate relationships from diverse datasets, aiding in forecasting outbreaks in regions like Peru. Such approaches offer valuable insights for public health authorities to implement timely preventive measures.

Climate-Based Dengue Outbreak Prediction with ML

Studies employing machine learning techniques, such as Support Vector Machines, have demonstrated the efficacy of analyzing climate variables like temperature, humidity, and rainfall in forecasting dengue outbreaks. These models provide timely information for health authorities to proactively implement preventive measures, thus mitigating the impact of dengue outbreaks.

Temporal Dynamics Analysis for Dengue Prediction

Temporal dynamics analysis involves a multi-stage machine learning approach to dissect the temporal relationship between temperature fluctuations and dengue occurrences. By integrating auto-encoding, data representation, and temporal clustering, these models offer accurate predictions of outbreak trends over time, providing crucial insights for public health planning and resource allocation.

Mathematical Models for Dengue Incidence Prediction

Machine learning techniques facilitate the development of mathematical models that predict dengue incidence based on weather data and population density. These models offer precise estimates of dengue cases, aiding in resource allocation and proactive public health planning.

Experiment Setup and Implementation

Our research focuses on developing a predictive model for dengue outbreaks in Sri Lanka, integrating machine learning techniques with microbiological data. By analyzing viral strains and genetic sequences, we aim to enhance the accuracy of outbreak predictions and identify potential risk factors associated with specific strains.

As our intial step, we have collected and preprocessed data from various sources, including environmental factors, historical dengue incidence rates. We have developed a data pipeline to streamline the data processing and feature extraction process, enabling us to generate comprehensive datasets for training and testing our models.

image info


Our project highlights the critical importance of incorporating microbiological data alongside traditional environmental factors in predicting and managing dengue outbreaks. Through the integration of genetic insights and machine learning techniques, we aim to enhance the accuracy of outbreak predictions and develop robust classification frameworks for discerning viral strains.

By adopting a multidisciplinary approach, we can significantly advance our ability to forecast and mitigate the impact of dengue outbreaks, not only in Sri Lanka but also globally. Our research underscores the necessity for further collaboration and exploration in this field to fully harness the potential of these approaches and develop targeted preventive measures.

Through the development of a predictive model that integrates both environmental and biological factors specific to Sri Lanka, we aspire to contribute to the ongoing efforts in combating dengue fever and ultimately reduce its burden on public health systems.