Forest Cover Prediction Using Machine Learning
Abstract
Forest land is highly required for developing ecosystem management. Any changes that occur in ecosystem should be carefully noticed to avoid further loss. This model is helpful in noticing the changes occurred due to heavy floods or any other calamities which affected the forest land. A machine learning Algorithm is used to predict the forest cover type using the cartographic variables. Kaggle dataset is taken for prediction of Forest cover type using Machine Learning. The complete dataset is composed by 581,012 instances, where each observation or instance corresponds to a 30 x 30 meters cell, for each observation, 12 measures (attributes) are given. The earlier models developed for prediction of Forest Cover type used cross fitting along with reduction of the variables used for training the dataset. This system splits the entire data into 10 equal groups. Our model processes the trained set collected from Kaggle records. An Exploratory Data Analysis (EDA) is done to make the data ready for application to applythe Machine Learning Algorithm. This will minimize the need to train the data every time we require testing for an instance.