Data Science Side Projects

Here include some of my work applying data analytics and modeling skills to real-time projects. Python and R are the 2 languages mainly used.

• In stock_timeSeries folder, I utilized ARIMA, LSTM, Prophet and additive model to play around yahoo finance stock price or sales data from open source. I also used LOWESS Regression fitting model on time series data points for forecast with 95% confidence interval.
• In statisticalAnalysis folder, I utilized Bayesian methodology with baseline uncertainty to predict 2024 US presidential election. Web scraping and regex techniques are used to access past 2020 election and most recent state-by-state polling data.
• In optimization folder, I did a few projects to find the optimal price, explore resource & inventory allocation.
  • Appraoches: KNN-Nearest Clustering, Bayesian Modeling (MCMC), Jupyter Widget, Feature Engingeer and Label Encoding
• In algoTrading folder, I played with different trading stratgies with regression, classification models to predict the likelihood and probability of price going up or down, and used backtesting to evaluate model and trading performance with different trading frequency. Then used Walk-Forward Validation to train and test a model.

Plase feel free to reach out (myLinkedIn) if you have any questions.