Weather Forecasting System
American Journal of Electronics & Communication · 2022
Throughout the ages and with the advancement of technology, weather has been one of the significant part of our lives.Though the main objective has always beenthe prediction of the figures that conveys us the facts about the weather, which includes temperature check, humidity, rainfall prediction, foresight of natural disasters, etc.The research paper mainly focuses on data science which is based on datasets and it validates the prediction of various atmospheric factor elements which makes it reliable to get the precise prediction.During the research, while the dataset were worked upon, few inconsistencies were encountered such as missing values in a dataset which were countered by using a pre-processing ML technique known as, "Label Encoding".The atmospheric factors used for weather prediction were: temperature, rainfall, wind speed, wind direction, humidity, etc.The classifiers used in our research were: Logistic Regression, Linear Regression, K-Neighbors Classifier, Random Forest Classifier, Decision Tree Classifier, Gaussian Naive Bayes Classifier (Gaussian NB), Multilayer perceptron (MLP), K-Means clustering, and Support Vector Classifier (SVC).During the research, the classifiers have derived different accuracies of prediction, with Random Forest andDecision Tree, and Decision Tree, both of which came up with the best results (i.e.1.0).