- M.S., Statistics and Machine Learning | Aix Marseille University | Eiffel Excellence Scholarship
- B.S., Economics, Quantitative Economics | National University of Colombia
Machine Learnig Engineer - Time Series Forecasting @ Orange
Data Scientist and Business Analyst @ Scotiabank Canada
Statistics and Econometrics Researcher @ Central Bank of Colombia
This is part of my research work at Orange. I used libraries for feature extraction such as tsfresh, Catch22, ROCKET, and TSFEL to forecast time series of multiple frequencies as a supervised machine learning task. During the process, I provided insights into the dependency between lags, the most effective libraries and features, and suitable prediction methods.
https://github.com/jdalonsos/Time_Series_Forecasting
Why do some smartphones cost more despite similar specs? Driven by curiosity, I analyzed smartphone pricing using data scraped from an e-commerce site. After rigorous cleaning and insightful visualizations, I uncovered the hidden influence of brand perception alongside technical specifications. This analysis revealed the true dynamics shaping smartphone market prices.
https://github.com/jdalonsos/Beyond-the-Specs-Decoding-Smartphone-Pricing-Strategies
In this notebook, we explore interpretability methods applied to a regression task on electric vehicle specifications. Using a dataset containing technical and performance attributes of electric cars, I aim to understand how different features—such as acceleration, top speed, and battery capacity—influence the predicted price of an electric vehicle. I employ both global and local interpretability techniques. Specifically, I use: SHAP values, Partial Dependence Plots (PDP), Individual Conditional Expectation (ICE), and Accumulated Local Effects (ALE), to shed light on the decision-making process of machine learning models such as Random Forest.
https://github.com/jdalonsos/InterpretabilityML
This project extracts real-time surf data from Surf-Report using Python web scraping techniques and visualizes the ideal surf conditions through an interactive dashboard built with R (Flexdashboard). It provides:
Live wave & wind data for a chosen location Key Performance Indicators (KPIs) like best surf time, highest wave of the week, and a surf quality rating User-friendly insights to help surfers optimize their sessions Tech stack: Python (BeautifulSoup, Pandas), R (Flexdashboard), Web Scraping, Data Visualization
Let me know if you want to tweak anything!
https://github.com/jdalonsos/jdalonsos/tree/main/assets/img/surf.png
