Main

 
Gerard circle crop 01 - small.png
 


GERARD BARBALICH

DATA SCIENTIST

Kia ora. I’m a Data Scientist in Poneke, Aotearoa (Wellington, New Zealand). I work with Harmonic Analytics currently, and previously with InternetNZ and Stats NZ.

I have a Masters in Science Communication and a Bachelors with Honours degree in Plant Biotechnology, both from the University of Otago. You can take a look at my LinkedIn profile here.

Here are some of my larger data science projects:

  • A supervised machine-learning classifier built for InternetNZ, which identifies fake online stores within .nz, and uses Airflow and Docker for automated scheduling - link. The structure of this supervised learning discrete classifier could be adapted for spam detection, fraud detection, or any other use case with a robust dataset of positive cases.

  • A ‘content-based filtering’ tool that recommends movies based upon customer preferences - link. This technique could be adapted to fit any situation where we want to make predictions about a customer, and have a robust history of data.

Here are some examples of classic data science tasks:

  • Clustering and visualising customer data using K-means - link.

  • Clustering text data for de-duplication using Levenshtein distance - link.

  • Analysing A/B tests with p-values - link.

Send me an email:

gerard [at] barbalich [dot] nz