Predicting the future (of equipment) using ML.NET

on YouTube ↗26m

" summary

Claim: predictive maintenance models are trainable end to end with ML.NET using CSV telemetry and run locally as an inferencing runtime. The demo uses ML.NET Model Builder and AutoML to try many trainers, produces a classifier, and scaffolds a console app that loads the .mlmodel and makes predictions; early trial accuracies quoted move from ~30% to ~50% on simple data. For much larger sets training time grows (example: 1 GB ≈ 3 hours), but small proofs run in minutes.

AI-written summary · about

" chapters (12)

1. 0:00Predictive equipment maintenance intro
2. 1:27Maintenance strategies evolution
3. 3:33ML concepts for developers
4. 6:25ML.NET features overview
5. 8:34Model Builder UI walkthrough
6. 10:39Demo setup and VS integration
7. 12:25Data classification project steps
8. 14:11Training models with AutoML
9. 16:40Model accuracy evaluation
10. 17:44Console app model integration
11. 19:53Key takeaways on ML.NET
12. 20:15Q&A

" related talks

• 45mMachine Learning for .NET developersOlia Gavrysh · NDC London 2020
• 59mMachine Learning simplified for Developers with ML.NETJernej Kavka · NDC Sydney 2020
• 59mMachine Learning in .NET (ML.NET)Agatha Stena · .NET Conf 2018
• 56mMachine Learning for .NET developersOlia Gavrysh · NDC Oslo 2020
• 46mMachine Learning for .NET DevelopersVeronika Kolesnikova · DevReach 2022
• 57mMachine Learning Made Easy – An Introduction to ML.NETAlexander Slotte · JetBrains .NET Day Online 2019
• 52mML.NET - Machine Learning made for .NETPeter Drougge · Update Conference Prague 2018
• 01h 03mMachine learning for C# developers: Introducing MLJeff Prosise · DotNext 2019 Moscow