/PredictCode

Standardised implementations of prediction techniques in the literature

Primary LanguageJupyter NotebookOtherNOASSERTION

Build Status

Predictive algorithms for crime

Python based implementations of algorithms to predict crime (which has a strong spatial component, e.g. Burglary). Current work has concentrated on giving implementations of algorithms from published sources.

  • Example Jupyter notebooks showing the different algorithms in action, together with discussion of the literature and implementation details.
  • Notebooks giving more technical details
  • Evaluation gives details about different evaluation/comparison methods for comparing predictions. (ToDo: Link to academic paper once finished.)
  • Documentation extracted from the doc strings in the Python code. (ToDo: Rebuild, as hasn't been compiled in ages).

GUI mode

Currently a work in progress. Can be run without installation via:

python -m open_cp

Install

Most notebooks can be run without installation. To install,

python setup.py install