Predicting Global Warming using Artificial Intelligence

Global Warming is often accountable for various phenomena, but many questions surrounding it are left unanswered. This paper discusses the solution to this issue by predicting global warming and its most influential causes through mathematical modeling. The initial steps of this project involved gathering credible data from sources such as the National Oceanic and Atmospheric Administration and the United Nations to begin data preprocessing. Before beginning the Exploratory Data Analysis, a hypothesis was formed: A linear regression model can be developed to predict the temperature anomaly. Correlation analysis was performed to eliminate multicollinearity by dropping redundant features and identifying the relationships between the target variable and eleven predictive variables. To proceed, it was verified that linear regression could be used on the current dataset. This involved ensuring data points were independent, and verifying homoskedasticity. The final step is developing the regression model to effectively predict features which determine temperature. This is achieved by solving for Y, or the dependent variable based on the values of the eleven independent features. The coefficients of the features are then computed based on the best-fit line. From the regression coefficients, it is evident that total greenhouse gas emissions, population growth, population density, and life expectancy significantly influence temperature anomalies. This research is important because it will aid people in making informed lifestyle changes by increasing awareness of how global warming can be mitigated.