Crime Prediction: 10-Wk Proof of Concept

Analytikus, LLC

This engagement delivers a working Proof of Concept for crime prediction.

Our experts use Power BI and Machine Learning to predict the probability of the occurrence of a crime, in a particular place, on a particular day and time.


Week 1-4

  • Analysis data sources
  • Transform data to fit the needs of our advanced analytical solution
  • Cleanse data and remove errors
  • Modify data types to support specific calculations

Week 5-7

Construction of prediction models to obtain a risk index by type of crime / hour / day / location. The following prediction techniques are used:

  • Neural networks
  • Logistic regression
  • Multinomial regression
  • Decision trees

Construction of explanatory early warning models for modus operandi. Types of modus operandi; up, to the low, news.

Week 8-10

  • Display of probability of occurrence by type of crime / day-time / modus operandi and location (block):

Map where you can choose:

  • The risk of occurrence indicator by type of crime / day / hour.
  • The granularity level of visualization: points, blocks, quadrants, zones, etc.
  • A particular polygon to filter the rest of the report

Context graphics where you can display the filter used on the map:

  • Risk factors
  • Historical view of crimes
  • Context variables


The following deliverables will be provided as part of this engagement:

  • Proof of Concept Summary and Recommendations
  • Dashboards in Power BI to analyze the results

At a glance