This notebook is an implementation of the work shown in the paper :

This repo is an implementation of the research paper "A Data Mining Approach to Predict Forest Fires using Meteorological Data." by P. Cortez and A. Morais. http://www.dsi.uminho.pt/~pcortez/fires.pdf

Using Regression to predict Forest Fires :: This is a regression task, where the aim is to predict the burned area of forest fires, in the northeast region of Portugal, by using meteorological and other data. Collected from UCI ML repository. Details can be found - (http://archive.ics.uci.edu/ml/datasets/Forest+Fires).

A comparision of four methods has been done ::

• Support Vector Regression
• Decision Trees
• Random Forest
• Deepnets(Keras with tensorflow backend)

Dataset used : https://github.com/geeky-bit/SVR_Decision-Trees_Random-Forests_DeepNeuralNets_to_PREDICT-FOREST-FIRES/blob/master/Datasets/forestfires.csv

Original dataset : http://archive.ics.uci.edu/ml/machine-learning-databases/forest-fires/

Note:: The codes are available freely. Feel free to add more codes. Spread the word!