Topic abandonned
Perform several Trainings into 1 solution and get each of them separately
Dear experts,
I am a new user of ML.NET and I am not clear about how to perform my needs.
My environment is the following :
windows 10, .NET Core 3.1 LTS and latest Visual Studio.
What I want is to train a ML to get 5 Labels form 1 Feature. From the tutorial I have seen the example of 1 Feature -> 1 Label training and I am not sure if it is possible to extend the Text classification for my case.
This is a simple example dataset of what I want to achieve :
1/ Feature = The cat lenght is 30cm
-> Animal=The cat / Extra= / Prop=lenght is / quantity=30 / unit=cm
2/ Feature = The dog density is 100g/cm3
-> Animal=The dog / Extra= / Prop=density is / quantity=100 / unit=g/cm3
3/ Feature = The horse number 2 speed is 45km/h
-> Animal=The horse / Extra=number 2 / Prop=speed is / quantity=45 / unit=km/h
1 Feature input -> 5 Labels to determine : Animal, Extra, Prop, quantity, unit.
I have a list of Feature and Labels into my existing dataset, then I want to apply the result of the training on whatever new Feature not in my dataset and find the 5 Labels above.
I am thinking about making 1 training for each Label in cascade.
A/ Train a ML to get Label=unit from Feature.
B/ Train a ML to get Label=quantity from Feature+unit.
C/ ...
and so on until I get the 5 Labels trainings.
I am not clear with 2 points :
- I have already done the 1st training and created the app into my solution.
I want to perform my second training into this solution. How can I do that ? - In the end I want to get the result of the 5 trainings on an independant Feature.
How can I access each trainings separately?
Best regards,
David.