PREDICTOR2: A MODEL FOR PREDICTING CO2 LEVELS IN THE ATMOSPHERE

Global warming stands as one of the most critical challenges of the modern era. Over
the decades, anthropogenic actions have collaborated with the climate changes suffered [Houghton
e Woodwell, 1989], including the rise in Greenhouse Gas (GHG) emissions, particularly CO2,
driven by fossil fuel combustion, among other sources. This study aims to utilize time-series data
on atmospheric CO2 levels, provided by NASA and monitored by NOAA, alongside AI models to
predict future indices. For this purpose, a custom Machine Learning (ML) model, PrediCtOr2, was
developed for this work, and its results were compared with those obtained using the Scikit-learn
library. The models were validated through the Least Squares Method (LSM). By the conclusion of
the experiment, the consistency of the models’ predictions was confirmed, validating their efficiency
for decision-making regarding this pressing issue.