We would solve the Forest Fire Prediction problem by using Machine Learning or Deep Learning where would estimate the percentage of fire within the forest using a dataset where we would determine coordinates of the hotspots where fire initialization has the most Probability by researching the feature which are crucial towards and we would determine the area of this widespread by selecting the features we would check and tells its importance in format of facts telling users the percentage we would determine that by changing the feature values which would effect our output value and note the change , percentage of predicting fire would come out from Normalizing the area output , we would use dataset from internet and customize the dataset adding new features from this dataset the output is area which we would determine it by svm or Neural Network while we store feature importance permenentaly and manipulate it according to the days probability of fire happening. We are thinking to use API for the days data and implying it into our Machine Learning Model. Further on we are trying to make Visualizer which would represent live Real Life Forest Condition and how would the fire spread and it's rate of fire spreading within the forest within a day.