Skip to content

zoelesv/datascienceportfolio

Repository files navigation

Data Science Portfolio

Welcome to my data science portfolio! I am a Data Scientist with 4+ years of experience in building data-driven solutions to solve complex business problems.


🔹 Featured Projects

1. Recipe Site Traffic Predictive Model

  • Goal: Classify high-traffic recipes to improve the product by increasing site traffic.
  • Tech Stack: Python, Logistic Regression, Random Forest, Feature Engineering, Data Normalization
  • Results: Achieved an AUROC of 0.902 with a Logistic Regression model and implemented a Key Performance Indicator (KPI) to maintain an 84% High Traffic Conversion Rate.
  • Details: Notebook

2. Color Analysis Neural Net

  • Goal: Classify facial images into their corresponding seasonal tones based on color theory.
  • Tech Stack: Python, Deep Learning, Neural Networks, Custom Triplet Loss, Computer Vision
  • Results: Achieved 74.47% accuracy on the initial dataset, with potential for improvement using a larger and more representative dataset.
  • Details: Notebook

3. Netflix Browsing Time Optimization

  • Goal: Minimize Netflix user’s browsing time by optimizing Match Score, Tile Size, Preview Length, and Preview Type.
  • Tech Stack: Python, A/B Testing, Factorial Design, F-tests, Partial F-tests
  • Results: Reduced browsing time to an average of 10.02 minutes with a 95% confidence interval.
  • Details: Notebook

4. E-commerce Cart Add-On Conversion Forecast

  • Goal: Predict the conversion rate for adding non-medical items to the cart and evaluate the potential impact of a proposed feature change on the e-commerce platform.
  • Tech Stack: SQL, Python, Time Series Analysis, Augmented Dickey-Fuller Test
  • Results: Forecast shows conversion rate reaching 20% in 3 months, with an upward trend correlated to revenue gains. The proposed feature change by the Product Team could accelerate this trend by increasing visibility of non-medical items during checkout.
  • Details: Report

🔹 Courses

  • Statistics - MSDS
  • Advanced Machine Learning - MSDS
  • Deep Learning - MSDS
  • Experiments in Data Science - MSDS
  • Time Series - MSDS

Feel free to reach out to me for collaboration or inquiries!

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published