This repository contains a handful of notebooks from my PhD research where I trained a machine learning model (i.e., the Self-Organizing Map) to identify the relationship between the largescale circulation over the North Pacific and heavy rainfall in Hawaii. Once this relationship was established, I then projected CMIP6 data onto the model to better understand how this relationship may change in a future climate.

In addition to the Self-Organizing Map, which is an unsupervised learning method, I also experimented with the K-Nearest Neighbors (KNN) algorithm, which is a supervised learning method.

The notebooks are read best in this order:

1. som_analysis_NPJ.ipynb
2. boxplots_cmip6.ipynb
3. disturbance_analysis_SOM.ipynb
4. plot_HI_rainfall_SOM.ipynb
5. KNN_NPJ.ipynb