Global warming brings negative impacts to people’s lives. In order to explore the factors that contribute to global warming and predict temperature changes, engineering models using Long Short-Term Memory and Extreme Gradient Boosting algorithms were built. With the global year-average temperature data after preprocessing by Python, the mathematical models were computed to discover the factors that influence the global temperature change with temperature parameters and fluctuations of anomalies, including the global average temperature, carbon dioxide concentration, solar activity, etc. With the multi-variable linear differential equation, we conclude that the global climate is determined by a variety of factors. The results implied that a high concentration of greenhouse gases had a very limited impact on climate change, and solar activity had the greatest impact on global temperature.