Probabilistic Forecasting Part 2 (Lag Features & NGBoost)

This is part two in a series of videos going over probabilistic forecasting.This is mostly a 'no-edit' video. I just let the camera roll as I worked on the script -- a live recording feel. This means the the video is longer, but if you watch the whole video you will learn something new. Code for this video is here: https://github.com/xvzftube/ts_vids/tree/main/forecast_m5_states Timecodes 0:00 Intro 0:07 Splash Screen 00:18 TODO Review (00:18) 03:15 Script Cleaning - Flake8 (03:15) 06:00 Flake8 Docs - Flake8 (06:00) 11:13 AleFixer - black (11:13) 14:11 Get entr running (14:11) 15:18 Look At Data (15:18) 17:30 Dig Into Script (17:30) 28:59 Grouped Lags, MA, EWM (28:59) 51:59 NGBoost Setup (51:59) 1:02:48 Play with NGBoost Object (1:02:48) 1:11:32 TODO Review (1:11:32) FREE AND OPEN SOURCE SOFTWARE THAT I USED FOR THIS VIDEO: 📽️ [LGPL] ffmpeg: https://ffmpeg.org 📃[Apache 2.0] NVIM: https://neovim.io/ 📦[GPL] Kdenlive: https://kdenlive.org/en/about/ #data-science #vim #nvim #python #forecasting #probabilistic-forecasting #machine-learning #ngboost #boosting #flake8 #black