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

GSVDplusplus model #40

Open
yustiks opened this issue Mar 30, 2021 · 5 comments
Open

GSVDplusplus model #40

yustiks opened this issue Mar 30, 2021 · 5 comments

Comments

@yustiks
Copy link

yustiks commented Mar 30, 2021

Hello!
I am trying to use the library in order to apply GSVD++ model with items side information, and after some time of training, I have the following prediction on a test data:

94      77755   nan
94      82092   nan
94      85906   nan
94      84238   nan
94      87209   nan
94      85765   nan
94      87759   nan
94      77166   nan

which means that method has found only nan values for users.
What could be the explanation?
PS: I run SVDpp model, and I have a precise predictions on the same data.

@yustiks
Copy link
Author

yustiks commented Mar 30, 2021

    return f(*args, **kwargs)
  File "/home/iivanova/miniconda3/envs/tf_gpu/lib/python3.7/site-packages/sklearn/metrics/_regression.py", line 183, in mean_absolute_error
    y_true, y_pred, multioutput)
  File "/home/iivanova/miniconda3/envs/tf_gpu/lib/python3.7/site-packages/sklearn/metrics/_regression.py", line 90, in _check_reg_targets
    y_pred = check_array(y_pred, ensure_2d=False, dtype=dtype)
  File "/home/iivanova/miniconda3/envs/tf_gpu/lib/python3.7/site-packages/sklearn/utils/validation.py", line 63, in inner_f
    return f(*args, **kwargs)
  File "/home/iivanova/miniconda3/envs/tf_gpu/lib/python3.7/site-packages/sklearn/utils/validation.py", line 664, in check_array
    allow_nan=force_all_finite == 'allow-nan')
  File "/home/iivanova/miniconda3/envs/tf_gpu/lib/python3.7/site-packages/sklearn/utils/validation.py", line 106, in _assert_all_finite
    msg_dtype if msg_dtype is not None else X.dtype)
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

that is the error

@giandos200
Copy link

Can you share the code? i think that some Hyper-parameter blow-up the rating!

@yustiks
Copy link
Author

yustiks commented Apr 1, 2021

Can you share the code? i think that some Hyper-parameter blow-up the rating!

this is my training data:
https://drive.google.com/file/d/1AlFV-NsildRNrJ7GxArxupexu5y1C4GM/view?usp=sharing

this is my testing data:
https://drive.google.com/file/d/1nUFtSXYrrloAB24vosrk1shLjzUhAJ3D/view?usp=sharing

side information of the items:
https://drive.google.com/file/d/1fX7NAQpEjLkXNfWyZvwMCxOmTjDtw5Ug/view?usp=sharing

@yustiks
Copy link
Author

yustiks commented Apr 1, 2021

@giandos200 Also, can you add the example of item-features matrix, please?

@yustiks
Copy link
Author

yustiks commented Apr 6, 2021

@eduardofressato Is there any examples on item-features matrix?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants