Angus Chan

Angus Chan

he/him

I am a PhD researcher at UCL Security and Crime Science, specializing in generating actionable insights through statistical modeling and machine learning in the security and crime domains. I am now seeking opportunities to apply my expertise as a data scientist.

Posts by Tags

Agent-based modeling

Agent-based modeling shoplifting simulation

less than 1 minute read

Published:

This project is part of the coursework from the simluation for research module. Utilizing Netlogo, the code creates a simulation in which security guard’s probaility of detecting a shoplifting and shoplifter’s probability of detecting a security guard nearby are manipulated. See more details at its repository shoplifting ABM

Crime Hamr Index

Crime Mapping

Customer Segmentation

DBSCAN (Density-Based Spatial Clustering of Applications with Noise)

Hotspot Policing

KMeans Clustering

Light GBM

Location- and demographic-based targeting business strategy

Logsitic Regression

Multinomial Naive Bayes

IMDB reviews sentiment classification

less than 1 minute read

Published:

This project is part of the coursework from the applied data science module. By training models from a large dataset containing 25k IMDB movie reviews that have been labeled as either positive or negative, it predicts another 25k movie reviews to be positive or negative. For more details, please refer to the corresponding repository IMDB_sentiment_classification.

Neural Network

Random Forest

Random Forest Classification

IMDB reviews sentiment classification

less than 1 minute read

Published:

This project is part of the coursework from the applied data science module. By training models from a large dataset containing 25k IMDB movie reviews that have been labeled as either positive or negative, it predicts another 25k movie reviews to be positive or negative. For more details, please refer to the corresponding repository IMDB_sentiment_classification.

Simulation

Agent-based modeling shoplifting simulation

less than 1 minute read

Published:

This project is part of the coursework from the simluation for research module. Utilizing Netlogo, the code creates a simulation in which security guard’s probaility of detecting a shoplifting and shoplifter’s probability of detecting a security guard nearby are manipulated. See more details at its repository shoplifting ABM

Support Vector Machines (linear SVC)

IMDB reviews sentiment classification

less than 1 minute read

Published:

This project is part of the coursework from the applied data science module. By training models from a large dataset containing 25k IMDB movie reviews that have been labeled as either positive or negative, it predicts another 25k movie reviews to be positive or negative. For more details, please refer to the corresponding repository IMDB_sentiment_classification.

Synthetic Minority Over-sampling Technique (SMOTE)

Time Series Forecasting

XGBoost