ANALYSIS OF WILDFIRE OCCURENCE IN AUSTRALIA USING DATA ANALYSIS TECHNIQUES

Abstract

Thousands of human lives are lost every year around the globe, apart from significant damage to property, animal life, etc., due to natural disasters. This project focused on Wildfire prediction. The work has been performed on building a predictive model for wildfires in Australia during the hottest period of the year. Datasets that have been used contain data of fire activities in Australia from 2005 to 2020. The work done for this project is divided into three parts: giving a brief description of algorithms and methods that will be used for predictive models, steps that will be followed for analyzing, preprocessing the data, and finally building the predictive model for Australian wildfires in December 2021. This project will also cover the topics of big data, deep learning and machine learning. Multiple steps will be followed in order to build the dataset. These steps include collecting an amount of data, using different preprocessing methods and techniques to correct data inconsistencies, and filtering the data used for the following process. Regarding the predictive models, multiple useful algorithms have been included that are being used for data mining, simulation, and testing.