I am a 22 years old Portuguese student with a Bachelor's in Economics and currently pursuing a Master's in Finance with a Major in Business Analytics. Having a background in Economics I have a major interest in the areas of education and development economics and data science has proven to be a very flexible tool to help in a wide range of initiatives designed to create impact. From tackling the education gap in less developed communities through data-based education as the Rlabs have done in South Africa, to mixing behavioral economics insights and data analysis to increase vaccination rates or decrease logging and deforestation.
Using a Passive-Aggressive Model to classify fake news. The model used is an "online" learning model that iteratively classifies a datapoint as "Real" or "Fake". If it gets it right, no adjustments are made, ergo the Passive, while if it gets it wrong, the decision weights are adjusted to get it righ. It is, in theory, a good model to classify a real time flow of data that is separable, that is, the classification line is well defined.
This was a project under the Data Analytics for Finance subject at Nova SBE. The objective was to evaluate the performance of a quantitative investment strategy using common performance evaluation techniques. The quality of the project was be based on the quality of the analysis, not the performance of the strategy. As such, a well-executed analysis of a strategy that performs poorly was perfectly acceptable.
In this project, under the Machine Learning subject at Nova SBE, we using a dataset from a large service provider of online advertising and digital marketing containing ads shown during 10 days we wanted to predict if a user will click or not click on an ad taking special attention to the role of the website, the position and display of the add and other features that can be controlled by the advertiser. Click Here for the presentation's pdf
In the final project of this subject we had to create a small website using React - Project DEMO - with interactive elements, simulating a the addition of Car trips and respective drivers. Alongside with the front-end, we had a back-end which stored the inputs while the server was running.
Data Analysis and Exploration using R and a Report summary of the exploration, answering specific questions and suggesting a possible redesign of the experince. This was done in the context of the subject Marketing Analytics during my Masters in Finance at Nova School of Business and Economics.
A project leveraging jupyter-notebook widgets and data visualization modules in order to facilitate initial data exploration. This was done in the context of working for the DSKC and it is only part of the iterations that led to the final project. However it is still fully functional and I use it on my day to day.
A quick data analysis on a merged AirBnB dataset of Lisbon & Porto listings aiming to predict, using only 3 features and as few visualizations as possible, if a listing would be available in the next 280 days or not. This a fun project specially when trying to choose what features to use and scaling down only to meaningfull visualizations.
This project consisted on scrapping the Top 100 IMDb movies and anwer questions using pandas and numpy in the context of the subject Data Curation during my Masters in Finance at Nova School of Business and Economics. I have made a personal addtion to the project providing useful visualizations for the different challenges.
This is a group project from my masters degree in Finance for the Data Visualization Course where we merged 5 World Happiness Report Datasets and performed a Data Analysis and visualization trying to extract trends and infer possible causality for different happiness factors.
Personal project to create a dashbord to check the evolution of my investment portfolio. It was also a way to develop my skills in Dash and Plotly libraries and well as some API data requests, connecting google Drive, google Sheets and Google Data Studio to have an updated portfolio anytime I ran the script.
