master-projects
Documenting a master's degree in data science
This repository collects a diverse set of projects developed through the course of my master's in data science at the Barcelona Graduate School of Economics (BGSE). Most of the projects were developed in Jupyter notebooks running Python 3.
- ML_toolbox: contains practical examples of machine learning models using scikit-learn. Additionally, a theoretical notebook on neural networks and a practical implementation using PyTorch are available.
- nlp: contains a several notebooks exploring the basics of natural language processing. Among these are: a 'manual' implementation of a text vectorizer, a sentiment analysis prediction and an exploration on word embeddings.
- reinforcement_learning: contains some very basic explorations of the field combining notebooks dealing with Markov processes and some others centered on Gaussian processes and Bayesian optimization.
- statistical_modeling: for the moment, the folder only contains a project exploring latent variable models and implementing the Expectation–maximization (EM) algorithm and the Gibbs sampler.
- dashboard_project: contains the code necessary for building a Dash interactive dashboard that contains data from New York city. In order to constantly update the dashboard and keep it running, I utilize several tools from Amazon Web Services (AWS).
- data_warehousing: simple projects that explore the process of collecting, storing and retrieving data. Includes some SQL code.