How Much Did it Rain? Predicting Real Rainfall Totals Based on Radar Data
This work addresses rainfall prediction for meteorology and related fields, but it is incremental as it applies existing methods to new data.
The researchers tackled the problem of predicting rainfall totals from radar data using machine learning models, achieving competitive performance in a global competition with a parametric k-nearest-neighbor model that took six days to compute.
We applied a variety of parametric and non-parametric machine learning models to predict the probability distribution of rainfall based on 1M training examples over a single year across several U.S. states. Our top performing model based on a squared loss objective was a cross-validated parametric k-nearest-neighbor predictor that took about six days to compute, and was competitive in a world-wide competition.