Adventures in Machine Learning

Distributed Time Series in Machine Learning - ML 088

Sep 22, 2022
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 2min
2
Databricks and Distributed Time Series Data
01:58 • 4min
3
Databricks Case Study - How to Develop a Time Series Forecast for KPI's
05:56 • 5min
4
How to Summarize a Time Series Forecast for a Distribution Center
10:37 • 3min
5
Is It Better to Summatively Sum Up Your Time Series Than Summarize Your Models?
14:00 • 2min
6
How to Consume the Real-World Feedback From Your Forecasts
15:37 • 5min
7
Kubernetes Distributed Python Training
20:21 • 2min
8
Scaling Up to 500,000 Forecasts on a Laptop
22:24 • 4min
9
Python Data Generation Is a Lot Faster Than Using NumPy
26:14 • 2min
10
Using Parallelized Libraries Can Save You Tons of Time and Tons of Performance
28:29 • 2min
11
ML Cool - How to Develop Models for Forecasts
30:49 • 6min
12
How to Get Super High Accuracy to 500,000 Forecasts
36:39 • 3min
13
Using Probability Estimation in a Visualization Is Important
39:24 • 3min
14
Using a Parallelized Library on the Cloud for Distributed Time Series Forecasts
41:55 • 5min