Posts tagged machine-learning.

8 posts · all tags

DateTitleTagsMark
2026-04-23 Three Ways to Look at Time
ST-ResNet decomposes crime patterns into three temporal scales and models each one separately. Clever architecture, but does it actually help with only four years of NZ data?
 data machine-learning open-data
2026-04-16 Teaching a Neural Network to Watch Crime Like Video
ConvLSTM was built for weather radar. Turns out predicting crime on a grid is basically the same problem. Here's how it works and what it learned.
 data machine-learning open-data
2026-04-09 Can You Beat Last Month?
Before deep learning gets a chance, we need to know how well stupid-simple models perform. Turns out, they put up a real fight.
 data machine-learning open-data
2026-04-02 What the Data Actually Shows
Before throwing deep learning at Auckland crime data, you need to actually look at it. Seasonal patterns, spatial hotspots, and the sparsity problem.
 data machine-learning open-data
2026-03-26 Crime as Video
Turn a million geo-tagged crime records into a 4D tensor by overlaying a 500m grid on Auckland. Crime prediction becomes video prediction.
 data machine-learning open-data
2026-03-26 Giving Crime a Place on the Map
Crime records come with names and codes but no coordinates. Here's how we joined 1.15 million records to Stats NZ boundary files and gave every crime a place in physical space.
 data machine-learning open-data
2026-03-26 Wrangling a Million Crime Records
NZ Police's crime dataset is publicly available, but it's UTF-16 encoded, full of trailing periods, and 32% of records don't know what time the crime happened. Here's how we cleaned it up.
 data machine-learning open-data
2026-03-24 Predicting Crime in Aotearoa
NZ Police publish over a million crime records openly. What happens when you point deep learning at them?
 data machine-learning open-data