This project applies smoothing techniques to temperature time series. I propose using a super-smoother with k-block cross-validation to overcome some of the significant difficulties that temperature data pose for other automated smoothing techniques.
This repository contains a short paper outlining how I arrived at the technique (Smoothing-Temperature-Data.pdf), the temperature data used, and a .rmd file with R code.
Adapted from my final project for the Fall '23 Computational Statistics course I took at Johns Hopkins for my MS.