dimsirovic/wiki_search

End-to-end search engine for Wikipedia articles

About Me

Hi, I'm Dinko!

If you like what you see here and want to chat, let's connect on LinkedIn 😄

My Projects

• 
  • Implemented an end-to-end scalable search engine for Wikipedia articles with MapReduce
  • Developed a RESTful API to rank results of user queries with PageRank and tf-idf analysis
  • Created a user interface with tunable weights

Tools: MapReduce, Flask, Jinja

• 
  • Demonstrated difference in distribution of 3 point field goal percentage for seeded versus non-seeded teams in March Madness
  • Justified use of permutational t-test with power analysis through Monte Carlo Simulations
  • Published Medium article demonstrating effective and clear communication of advanced statistical concepts

Tools: dplyr, ggplot2

In Progress

• College Football Game Prediction | Python, SQL
  • Goal: Collect data manually from the API at collegefootballdata.com ✔️
  • Goal: Predict winner of a college football game given season average statistics, talent, and returning production ✔️
  • Goal: Deploy model, allowing users to select any 2 teams since 2015 to predict winner of hypothetical matchup

Tools: Web Scraping, Pandas, Scikit-Learn

Coming Soon!

These machine learning projects are on the docket and I'm very excited to share them upon completion

• Customer Churn | R
  • Goal: Predict if a bank customer will turnover on their credit card services given card type; marriage, income, and education status, etc.

Tools: dplyr, ggplot2, tidymodels

• Spotify Popularity | Python
  • Goal: Predict how popular a song will become on Spotify given characteristics such as danceability, energy, key, etc.

Tools: Pandas, Scikit-learn