Forest-Fire-Classification-

The dataset is designed for binary classification of Fire and No-Fire detection in the forests landscape. You should do this classification using logistic regression. Our goal is to complete the following steps: a) Pre-processing: The size of the images may be different. Resize them. Then normalize the images. Determine the target of each image. (using glob and OpenCV libraries). b) Split the data into train and test. c) Run logistic regression on train data. d) Report accuracy and confusion matrix for test data. e) Find the best probability threshold for the training data and report the accuracy and confusion matrix again. f) Save the best model. import this model to another file (e.g., classifier.py). Download some forest images (fire and non-fire). feed this image to your classifier and make a prediction. Finally, show the label of each image with its probability in the photo. If the photo was fire, with red color, otherwise green color (similar to below)