adidam1399/Algerian-Forest-Fires-Prediction-Time-series-analysis-classification

Different supervised learning classifiers trained on Algerian Forest Fires dataset to predict imminent Forest Fires. # Algerian-Forest-Fires-Prediction-Time-series-analysis-classification Different supervised learning classifiers trained on Algerian Forest Fires dataset to predict imminent Forest Fires. The dataset is Algerian Forest Fires Dataset - which is a time series dataset. Used to predict whether forest fire occurs given certain features in the imminent future. Different supervised learning classifiers such as Logistic Regression, Perceptron, K-Nearest Neighbors, Random Forests, Naïve Bayes Classifier, Support ## Vector Machines (SVMs), Neural Network architectures are trained on Algerian Forest Fires dataset to predict imminent Forest Fires. Feature selection for models is done using Principal Components Analysis (PCA) & Sequential Feature Selection (SFS) on training data. Achieved an accuracy of 89 % on test set using SVM classifier with Linear Kernel trained on features selected using PCA