Abstract
Weather forecasting through neural networks has increased and shown the potential for greater accuracy over recent years. Among numerous techniques, machine learning models provide more precise weather and climate prediction outcomes. The objective of this research was to analyze the highest and lowest monthly temperatures, as well as the highest wind speeds, in selected Nigerian cities, including Abuja, Lagos, Sokoto, Maiduguri, Calabar, and Port Harcourt through the use of cutting-edge machine learning technology such as deep learning (DL), and Convolution Neural Network (CNN). Our research approach involved compiling data on maximum and minimum temperatures and wind speeds from specific cities in Nigeria every month from 2000 to 2023. By successfully utilizing AMI, we pinpointed the optimal variables necessary for precisely evaluating the six cities as we built our model. The CNN algorithm stood out as a top-tier model in the test results due to its precise estimation of city temperature and wind speed values, highlighting exceptional generalization ability and minimal variance compared to the DL model.