Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- Intel Community
- Software
- Software Development SDKs and Libraries
- Intel® oneAPI Data Analytics Library
- Linear regression module, issues

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

Reinhard_M_

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

09-23-2017
08:16 AM

124 Views

Linear regression module, issues

I have examined the linear regression module of the DAAL 2018 (Windows version)

1) In the "single_beta" submodule, the RMS error and the variance are inconsistent, i.e. var(j) != rms^{2}(j)*n/(m-p-1)

IMO, in linear_regression_single_beta_dense_default_batch_impl.i, line #316 should be

pRms

instead of

pRms

2) The "group_of_betas" submodule contains an unusual definition of the goodness parameter R^{2}

To my best knowledge, R^{2} runs from 0 ("no fit") to 1 ("perfect fit"). In the DAAL implementation (and documentation),

R^{2} runs from 0 ("no fit") to 1/n ("perfect fit"). I don't whether this is a buf or a feature

Link Copied

5 Replies

Shaojuan_Z_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

09-27-2017
03:02 PM

124 Views

Hi Reinhard, which file are you referring to? thanks

Reinhard_M_

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

09-28-2017
05:35 AM

124 Views

Topic 1: The file is "linear_regression_single_beta_dense_default_batch_impl.i", line #316

Topic 2: IMO, the regression sum of squares RegSS is just the squares and not the mean square. Therefore,

RegSS != TSS - ResSS;

In your example, TSS-ResSS ≈ n*RegSS. I think, this is the root cause if the R^2 issue.

Shaojuan_Z_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

09-28-2017
10:16 AM

124 Views

Shaojuan_Z_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

09-29-2017
11:57 AM

124 Views

Hi Reinhard, we are analyzing the question and will get back to you with details. Thanks!

VictoriyaS_F_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

10-02-2017
06:50 AM

124 Views

Hello Reinhard,

Both of your observations are correct –there are bugs in the quality metrics for linear regression in Intel DAAL 2018 Gold. We will fix them in one of future releases of the library.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

For more complete information about compiler optimizations, see our Optimization Notice.