This is the repository for the final project of the course Bayesian Learning and Montecarlo Simulation at Polimi.
Human emissions of carbon dioxide and other greenhouse gases are a primary driver of climate change and present one of the world’s most pressing challenges. To understand this complex issue better our analysis will focus on:
-
Long-Term Trends: we will examine how CO2 emissions have evolved over the past century, identifying key periods of increase or decrease in the speed of emissions and correlating these with historical events.
-
Recent Factors Influencing CO2 Emissions: we will analyze the impact of various factors on CO2 emissions from 2006 to 2009. This includes investigating the relationship between CO2 emissions and variables such as energy consumption, GDP and the adoption of low-carbon energy sources.
Brief description of the datasets:
-
Dataset 1 (Long Term) encompasses annual CO2 emissions for each year from 1900 to 2022. This dataset provides a comprehensive perspective on CO2 emissions over more than a century, allowing us to identify long-term trends and patterns.
-
Dataset 2 (Recent Factors) is more granular and concentrates on the years 2006 to 2009. It includes various features such as energy use per capita, GDP, population, CO2 emissions per capita, the percentage of low-carbon energy in total energy production, urbanization levels, and internet usage. This dataset allows for a detailed analysis of specific factors that may influence CO2 emissions during a recent, focused timeframe.
By integrating these two datasets, we aim to analyse CO2 emissions over time and identify the factors driving changes in recent years.
- Change point model on mean and variance
- Bayesian AR(1)
- White noise model with switching variance
- Spike-and-slab prior
- Bayesian linear regressions