Machine Learning Approaches to Addressing Climate Change Health Impacts

Abstract

With far-reaching effects, climate change is a pressing global issue. To make educated decisions and work to lessen the effects of climate change, it is essential to accurately predict future trends. Throughout human history, weather and climate forecasting have been essential for enabling efficient agricultural planning, protecting against natural disasters, and facilitating strategic decision-making across a variety of sectors. In this regard, it is imperative that forecasts be accurate and timely, and machine learning holds promise for increasing the precision and speed of prediction. This study uses online data from 2000 to 2023 to estimate future climate fluctuations using machine learning techniques, specifically advanced learning algorithms. I collect my data from Kaggle, and this dataset consists 48,000 data and 13 features have been used in this dataset. 'Date', 'Country', 'Temperature Anomaly (°C)', 'CO2 Level (ppm)', 'Extreme Weather Event', 'Economic Impact (USD)', 'Population Affected', 'Hour', 'NowCast Conc.' Etc. are features. Maintaining indoor air quality requires regular forecasting and monitoring of air pollution. As a result, machine learning (ML) has demonstrated potential in surpassing conventional methods in the prediction of the air quality index (AQI). The condition of the atmosphere is gauged by the air quality index (AQI). It determines how short-term health effects of modest exposure will manifest. Public education on the harmful effects of ambient pollutants on health is the aim of the AQI. After gathering the data and processing it all, we created a processed dataset. Using the previously processed dataset, we applied machine learning techniques. Using all 13 features, analysis between of all the related features. We use naïve Bayes, XG Boost, decision trees, random forests, and support vector machines (SVM). Applying 3 method without sampling, random under-sampling and SMOTE techniques among models outperformed the other five algorithms in our experiment in terms of accuracy; the random forest classifier's accuracy was 99.72%. AQI has 6 categories for apply the model.