Hack the North 2017 - Machine Learning on Calgary Fire Dept Data
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
Response for Fire.ipynb


Hack the Fire

Hack the Fire uses Calgary Open Data as training data so it can predict based only on date and # of incidents whether a response by the Calgary Fire Department was to a Fire or to another incident (health, misc...).

It uses a primitive Perceptron Machine Learning algorithm to build a predictive weighting.

With many loops of the algorithm, the weighting becomes more accurate reaching 95% accuracy after 500 iterations.

New test data can then be manually input to test the predictive capabilities.

Data Source

Tech Stack

  • Python 3.6
  • Pandas
  • NumPy
  • Jupyter

Built at Hack the North 2017. Not under ongoing development.