Coronavirus disease 2019 (also known as COVID-19) is a vastly infectious virus instigated by the coronavirus-2, which causes severe acute respiratory illness (SARS-Cov-2). Scientists and researchers are conducting a number of studies to better understand the COVID-19 pandemic's behavioral nature and spread, and machine learning provides useful tools. We used machine learning techniques to study the effect of climate conditions on daily instances of COVID-19 in this study. The study has three main objectives: first, to investigate the most climatic features that could affect the spread of novel COVID-19 cases; second, to assess the influence of government strategies on COVID-19 using our dataset; third, to do a comparative analysis of two different machine learning models, and develop a model to predict accurate response to the most features on COVID-19 spread. The goal of this research is to assist health-care facilities and governments with planning and decision-making. The study compared random forest and artificial neural network models for analysis. In addition, feature importance among the independent variables (climate variables) were identified with the random forest. The study used publicly available datasets of COVID-19 cases from the World Health Organization and climate variables from National Aeronautics and Space Administration websites respectively. Our results showed that relative humidity and solar had significant impact as a feature of weather variables on COVID-19 recorded cases; and that random forest predicted accurate response to the most climatic features on COVID-19 spread. Based on this, we propose the random forest model to predict COVID-19 cases using weather variables.