Skip to content
View beckerchTRJ's full-sized avatar

Achievements

Achievement: Quickdraw

Achievements

Achievement: Quickdraw

Block or report beckerchTRJ

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
beckerchTRJ/README.md

hi, i'm christian.

data science + economics, based in austin. usc '25 — ms in applied data science, bs in economics & data science.

selected work

project what it is tools
Can a model beat a prediction market? Six strategies, three months of simulated trading, and a 44,135-price accuracy study, judged by rules written in advance. Answer: no, and the repo shows why. Python, pandas, SQLite
Who can walk to transit in LA? 39 agencies' schedules plus Census blocks: 81% of LA County lives near a bus stop, 10% near rail, and 20 well-placed stations would reach 930,000 more people. GeoPandas, H3, D3.js
Yelp rating prediction 1st of 110 in a recommender-system competition. RMSE 0.9747 under Spark RDD-only and strict runtime limits; collaborative filtering made it worse, feature engineering won. PySpark, XGBoost
Libraries and graduation rates Ridge and lasso regression across 3,218 U.S. counties. Poverty dominates, but library circulation survives regularization. scikit-learn, pandas, Plotly
Bias in comedy scripts Bias-type classifiers (SVM → BiLSTM → fine-tuned BERT at 95% accuracy) applied to 21 film scripts. PyTorch, Transformers
Sharded retail database Central + sharded MySQL design with a desktop admin tool and a web customer search. MySQL, Flask, Tkinter

toolbox

  • languages: Python, SQL, C++
  • analysis & ml: pandas, NumPy, SciPy, scikit-learn, XGBoost, PyTorch, Transformers
  • geospatial: GeoPandas, H3
  • data: Spark, MySQL, MongoDB, Firebase
  • communication: Matplotlib, Seaborn, Plotly, D3.js, Power BI

also

  • 🌱 currently reading Statistical Rethinking
  • 🏕 i love camping
  • 🐈‍⬛ my cat, batman, is pictured in my profile pic helping me code
  • 📫 beckerch@usc.edu

Pinned Loading

  1. ECON-Library-Project ECON-Library-Project Public

    Do public libraries move graduation rates? Ridge/lasso regression across 3,218 U.S. counties using ACS, PLS, and SAIPE data

    HTML 1

  2. nlp-bias-in-comedies nlp-bias-in-comedies Public

    Classifying bias types in comedy film scripts: SVM, Perceptron, BiLSTM, and fine-tuned BERT (95% accuracy)

    Jupyter Notebook

  3. Yelp-Rec-Competition Yelp-Rec-Competition Public

    1st of 110: Yelp star-rating prediction with engineered features + XGBoost under Spark RDD-only constraints (RMSE 0.9747)

    Python