Skip to content
View sidneykung's full-sized avatar

Block or report sidneykung

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.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

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

Report abuse
sidneykung/README.md

Welcome 👋🏼

I’m a Data Scientist based in New York with a passion for understanding people through data. Leveraging my background in public relations and crisis management, I help data-driven companies tell their stories intuitively and meaningfully. I love chatting about machine learning, data privacy ethics and the attention economy. In my spare time, you can catch me searching for the best restaurant in the city.

Let's Connect!

LinkedIn Tableau Medium Twitter Gmail

Projects

This project automates Twitter hate speech detection with classification modeling. To prepare the data, 24,802 tweets were preprocessed using NLTK and RegEx. The final model was a Logistic Regression classifier that used CountVectorizer for feature engineering. It achieved a Recall (TPR) of 0.624 and an interactive version has been deployed on Heroku.

Time series modeling project to forecast LA reported crime rates based on 10 years worth of data. The final SARIMA model predicted the monthly average crime count, with an RMSE of 24.66. Exploratory phase discovered that the top 3 high-risk geographic areas in LA are 77th Street, Southwest and North Hollywood, with Latinx and Black victims being disproportionately impacted.

Binary classification project to predict whether oro not a client will default on their credit card. Baseline models included K Nearest Neighbors, Logistic Regression and Decision Tree models. By using GridSearchCV, the tuned Random Forest model was optimized and reached an F1 score of 0.5412.

This regression analysis project predicts house prices in King County Seattle, Washington. The dataset was sourced from Kaggle and included 17,290 rows of real estate data. Each iteration used feature selection methods to optimize each model’s RMSE. The final Linear Regression model achieved an RMSE of 214,529.87.

Published Articles

Towards Data Science

The Startup

Pinned Loading

  1. twitter_hate_speech_detection twitter_hate_speech_detection Public

    Capstone project to automate Twitter hate speech detection with classification modeling.

    Jupyter Notebook 28 13

  2. LA_crime_forecasting LA_crime_forecasting Public

    Time series modeling project to forecast LA reported crime rates based on 10 years worth of recent data.

    Jupyter Notebook 1 3

  3. cc_default_prediction cc_default_prediction Public

    Classification project to predict whether a person will default on their credit card or not.

    Jupyter Notebook 3

  4. Housing_Price_Model Housing_Price_Model Public

    Project that builds a model to predict house prices in King County Seattle, Washington.

    Jupyter Notebook