[Model] Implement Neural Collaborative Filtering with MXNet

[xinyu-intel]

Description

Implement Neural Collaborative Filtering with MXNet according to mlperf v0.5 configuration.

Evaluation accuracy on MovieLens 20m dataset:

dtype	HR@10	NDCG@10
float32	0.6393	0.3849
int8	0.6366	0.3824

@pengzhao-intel @TaoLv @zhreshold

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage:
• Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
• Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
• Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
• Code is well-documented:
• For user-facing API changes, API doc string has been updated.
• For new C++ functions in header files, their functionalities and arguments are documented.
• For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
• Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html
• To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

• data preprocess for MovieLens datasets
• ncf training on MovieLens dataset
• ncf evaluation and benchmark on MovieLens dataset
• ncf int8 quantization on MovieLens dataset
• ncf training/inference/quantization smoke test

Comments

• If this change is a backward incompatible change, why must this change be made.
• Interesting edge cases to note here

[pengzhao-intel]

@wkcn could you help take a review in case you're interested in?

[codecov-io]

Codecov Report

Merging #16689 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #16689      +/-   ##
==========================================
- Coverage   67.06%   67.05%   -0.01%     
==========================================
  Files         268      268              
  Lines       29748    29748              
  Branches     4398     4398              
==========================================
- Hits        19951    19949       -2     
- Misses       8536     8537       +1     
- Partials     1261     1262       +1

Impacted Files	Coverage Δ
python/mxnet/image/image.py	75.83% <0%> (-0.38%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ef19b09...a134631. Read the comment docs.

[wkcn]

@pengzhao-intel Sorry that I am not familiar with neural collaborative filtering, but I try to learn it : )

[pengzhao-intel]

@xinyu-intel is this work done and ready to merge?

[xinyu-intel]

yes:)

[wkcn]

Thank @xinyu-intel for the great contribution!
Sorry that I am busy recently, and I am taking a review on this PR today.

[pengzhao-intel]

LGTM

Merging now and we can continuously improve the case :)