CSSP China ML: Predicting weather in North East China using machine learning

This repo contains demonstrations of the techniques developed to predict temperature and precipitation in North East China using climate index timeseries data for the purpose of maize yield forecasting. These demonstrations were developed using R and Python, and can be interactively explored online using the Binder instance setup in this repository.

Using the Binder

Click the badges above to launch a live, interactive Jupyter Lab or RStudio session in your browser. Execute the notebooks/scripts, edit them, create new ones and upload your own data to analyse.

You can also launch each of the analyses in this repo from the Jupyter Lab file browser, or select RStudio from the launcher. You can also launch them from the badges in the sections below.

The sessions last for as long as you are using them and will disappear once you are inactive for 10 minutes. Any unsaved progress will disappear (each session starts with exactly what is in this repo), so if you want to save any progress/edits you make you will need to download them to your computer.

For more information on this Binder setup please see informatics-lab/binder_rstudio_jupyterlab_example.

1. Feature selection: Random Forest (R)

Things

2. Interpretation: Clustering, PCA, Causal & Bayesian networks, Dynamics plots (R + Python)

Stuff

3. Predictive model: Multivariant linear model(?)

Bits

References

EXAMPLE BADGES