Andres Zarta

Hello! My name is Andrés Zarta and I'm a Machine Learning Engineer from Bogotá, Colombia. Both at work and on my spare time I like to sharpen my Python, ML and Deep Learning skills so I can contribute to cool projects. I like to use Python for Data Science, Tensorflow and Trax for Deep Learning, Docker/Linux for deployment and Go for everything else. When I'm far from a computer I like to go hiking.

I got my Master's Degree in Operations Research back in 2016. My thesis centered around the generation of alternative rock basslines given a melody and a drumbeat using metaheuristics and optimization techniques. The paper can be found here if you'd like to check it out.

Latest Work

Q Learning using Numpy for Taxi V2 - Python

Straightforward Q Learning Implementation for OpenAI Gym's Taxi-V2 using plain Numpy ![](public/taxi.gif) __Tags__: Reinforcement Learning, Q Learning, OpenAI Gym, Numpy

Forest Cover Type Classification - Jupyter Notebook

Two separate Jupyter Notebooks showcasing exploratory analysis and predictive modelling for the Forest Cover Type dataset using Random Forest (because forests, bad joke, I know) and Artifical Neural Networks. __Tags__: Multiclass Classification, Exploratory Analysis, Random Forest, Multilayer Perceptron, PCA

Indica REST API - Python, PostgreSQL

REST API for a seed exchange app developed using Flask and PostgreSQL __Tags__: Flask, PostgreSQL, REST API, seed exchange

Dogs vs Cats Image Classifier - Jupyter Notebook

Notebook showcasing different strategies to develop a convnet based classifier that distinguishes dog images from cat images.

Tags: Computer Vision, Convolutional Neural Networks, Data Augmentation, VGG16 ,Tensorflow, Keras

Book Cover Reader - Python

Take picture of a book cover and let Google use their convnet fueled Vision to capture the text within. Find the corresponding book in the Books API and log the most relevant info (and errors) to BigQuery.

Thanks to Google for powering the book’s inventory management of Tu Libro a Ciegas!

Tags: Google Books API, Google Cloud Platform: [Storage, Vision, Bigquery]

Pulsar Star Detection - Jupyter Notebook

Can we predict a pulsar star in the universe? Let’s use the HTRU2 dataset and ML techniques to find out!

Tags: Binary Classification, XGBoost, KNN, NB, Logit, Exploratory Analysis, Outlier Analysis, Dimension Reduction, Classifier Comparison

Mushroom Classification - Jupyter Notebook

Is a mushroom safe to eat or deadly poison?

Tags: Binary Classification, Exploratory Analysis, Feature Importance, Feature Interpretation, XGBoost, SVC, NB, Logit

Get In Touch!

Send me an email. I’m also on Github and Linkedin.