dotnet / dotnet/machinelearning
Question: Realising own algorithms
Open
area-Core
untriaged
- Dominant language
- C#
- Stars
- 9.4k
- Forks
- 2k
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 11
Description
### Question
Is it possible with ML.NET/Microsoft.ML to realise an own algorithm?
We tried different approches like CustomMapping, ITransformer, trainer to realised it but at some point we got into trouble with internal properties/classes. Especially saving a model went down because of the need of "meta" data which we couldnt write because of internal properties. We tried reflection on some of the properties but loading went down then. So is there any possibilies yet to realise a own custom solution?
Contributor guide
Assessment
This issue has not been assessed yet.