Data project: Using machine learning with multiple multi class classifiers to correctly classify SafeGraph Patterns data as either airports, bus stops, or train stations using foot traffic, dwell time, visitors, & popularity

Hello all, I’ve recently been working on a project using Safegraph patterns data. The project revolves around POI classification using various Machine Learning models and provides insights regarding the utility of Safegraph data in the realms of Machine Learning. The various classification algorithms used in this project look to correctly classify records are either Airports, Bus stops, or Train stations using foot traffic, dwell time, visitors, and popularity metrics, and the results are very interesting.

Here is the link to the notebook file for the project: Google Colab

This is the first in a series of notebooks regarding this particular problem statement. There is more to come, such as model tuning through cross-validation and PCA (notebook 2) and Classification using Deep Learning through the Spark Multi-Layer Perceptron Classifier (notebook 3).

Let me know if you have any questions.