In this module, we will delve into fundamental concepts in machine learning. These concepts will serve as the foundation for our deep learning section of the course. we do this module in 4 parts:
1- Data preparation and fundamentals review
2- Decision tree based models for timeseries forecasting (this video)
3- Timeseries challenges
4- ML timeseries forecasting in Python
Timelines:
0:00:00 road map and recap
0:01:36 How decision trees work for timeseries - the fundamental questions
0:05:12 Decision tree regression (intuitive example)
0:17:11 Decision criteria for a timeseries regression task. what feature to start with and where to pu the cut off?
0:22:15 How to sample the data for splitting in a decision tree? Naive, histogram and GOSS
0:26:44 How to split the samples? Greedy vs Non-Greedy
0:29:20 How to grow a tree? Depth-wise, Leaf-wise or symmetric
0:33:30 How to combine trees? bagging vs boosting
0:36:34 putting it together: DT, RF, XGboost, Catboost and LightGBM
0:40:30 Going over a simple example in Python (deep dive into intuition)
0:56:23 Forecasting into the future and challenges of machine learning modeling for timeseries
Relevant playlists:
Deep Forecasting Concepts, simply explained: https://www.youtube.com/playlist?list=PL2GWo47BFyUPW_lptTNwpKNrpEQvUZerR
Machine Learning Codes and Concepts: https://youtube.com/playlist?list=PL2GWo47BFyUNeLIH127rVovSqKFm1rk07&si=lCPyHenEQYBCJzQ_
Deep Learning Concepts, simply explained: https://www.youtube.com/playlist?list=PL2GWo47BFyUO6Fiy2mJCxR8sUrBEfT6BM
Instructor: Pedram Jahangiry
All of the slides and notebooks used in this series are available on my GitHub page, so you can follow along and experiment with the code on your own.
https://github.com/PJalgotrader