Demystifying Gaussian Processes by Building One from Scratch with CO2 Data
- Introduction to Gaussian Processes (GPs) - flexible machine learning models that estimate uncertainty
- Implementing GPs from scratch in NumPy to gain deeper intuition
- Using monthly atmospheric CO2 concentration data from Mauna Loa observatory as an example
- Separating data into linear trend, seasonal trend, and residual components
- Linking to GitHub repo with full NumPy implementation of GPs