Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Best algorithm to use (ML.Net 1.4.0) #4502

Closed
hahsm opened this issue Nov 25, 2019 · 2 comments

Comments

@hahsm
Copy link

@hahsm hahsm commented Nov 25, 2019

Hi,
I have a dataset in the following format,

Item (feature) | Model (feature) | Brand (feature) | ItemId (PredictedLabel)
Iphone X | X MAX | Apple | 1
Samsung 10S | 10S | Samsung | 2
Samsung Note 9 | 9 | Samsung | 3
Likewise..

Based on the Item, Model and Size given as test data, I wanted to predict the ItemId. How is this possible, and what algorithm should I use to achieve this. I tried using Multiclass classification, and since Label column contains unique sequential values, each is considered as a separate class, and even with few data, it takes a very long time to train, with 0% accuracy. I am using ML.Net 1.4.0. Would appreciate if a sample code is shared.

Thank you.

@codemzs

This comment has been minimized.

Copy link
Member

@codemzs codemzs commented Nov 25, 2019

Please consider looking at our samples that explain how to use multi-class trainers. Also look at samples that show how to use transforms.

There is no such thing as best algorithm because it really depends on your dataset and problem that you are trying to solve. I will recommend building a basic pipeline from the samples just to get something working and then playing with the hyperparamters/trainers. Also please consider using more examples and may use of our sample datasets used in the samples.

@codemzs codemzs closed this Nov 25, 2019
@justinormont

This comment has been minimized.

Copy link
Member

@justinormont justinormont commented Nov 25, 2019

For the unique label values, see reply here:
#4473 (comment)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
3 participants
You can’t perform that action at this time.