hurricane-prediction

Jupyter notebooks with tools to analyze and predict the paths of hurricanes using historical hurricane paths and global topology data

This uses a topography dataset as well as a historical hurricane dataset. The idea is to use topography data alongside historical hurricane information (paths and airspeeds) to train an agent capable of predicting future location and max airspeed given the current data on a hurricane.

Datasets:

1. NOAA 2-Minute Gridded Global Relief Data

2. NOAA Hurricanes and Typhoons, 1851-2014

Requirements:

python 3.6.0
h5py (2.7.0)
numpy (1.13.1)
matplotlib (2.0.2)

Instructions:

1. extract datasets.7z into the folder ./datasets
2. run notebooks

TODO

• Incorporate satellite imagery to help learn climate data
• Incorporate temperature data
• actually build a model, ... yeah, ...