${\text{DS}}^2$

Data Science with Dimitrije Stankovic

Welcome to ${\text{DS}}^2$, a place to learn more about Data Science and any adjacent topics that interest me at any given time. I just finished my master's thesis on the prediction of electricity demand, so my first series will be a deep dive into forecasting. Hopefully you find it useful too!

Traditional Forecasting 101

• Prelude to Traditional Forecasting Methods
• Univariate Regression (From AR to SARIMA)
• Basic Model Evaluation and Selection
• Information Criteria and Grid Search
• Multivariate Regression (Adding an X)

ML Forecasting 101

• Neural Networks

• RNN for Numerical Forecasting

  • Vanilla RNN and the Vanishing Gradient Problem

  • LSTM

  • GRU