Scalable Learning and MAP Inference for Nonsymmetric Determinantal Point Processes
Authors: Mike Gartrell, Insu Han, Elvis Dohmatob, Jennifer Gillenwater, Victor-Emmanuel Brunel
ICLR 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Through evaluation on real-world datasets, we show that our algorithms scale significantly better, and can match the predictive performance of prior work. We combine the above contributions through experiments that involve learning NDPP kernels and applying MAP inference to these kernels to do subset selection for several real-world datasets. These experiments demonstrate that our algorithms are much more scalable, and that the new kernel decomposition matches the predictive performance of the decomposition from prior work. |
| Researcher Affiliation | Collaboration | Mike Gartrell Criteo AI Lab m.gartrell@criteo.com Insu Han KAIST insu.han@kaist.ac.kr Elvis Dohmatob Criteo AI Lab e.dohmatob@criteo.com Jennifer Gillenwater Google Research jengi@google.com Victor-Emmanuel Brunel ENSAE Paris Tech victor.emmanuel.brunel@ensae.fr |
| Pseudocode | Yes | Algorithm 1 Greedy MAP inference/conditioning for low-rank NDPPs |
| Open Source Code | Yes | Code for all experiments is available at https://github.com/cgartrel/scalable-nonsymmetric-DPPs. |
| Open Datasets | Yes | 1. Amazon Baby Registries: This dataset consists of registries or |
| Dataset Splits | Yes | We use a small held-out validation set, consisting of 300 randomly-selected baskets, for tracking convergence during training and for tuning hyperparameters. A random selection of 2000 of the remaining baskets are used for testing, and the rest are used for training. |
| Hardware Specification | No | The paper does not explicitly mention specific hardware details such as GPU models, CPU models, or memory specifications used for running the experiments. |
| Software Dependencies | No | We use PyTorch with Adam (Kingma & Ba, 2015) for optimization. The paper mentions PyTorch and Adam but does not specify their version numbers. |
| Experiment Setup | Yes | We use a small held-out validation set, consisting of 300 randomly-selected baskets, for tracking convergence during training and for tuning hyperparameters. ... We use PyTorch with Adam (Kingma & Ba, 2015) for optimization. We initialize C from the standard Gaussian distribution with mean 0 and variance 1, and B (which we set equal to V) is initialized from the uniform(0, 1) distribution. ... See Appendix B for the hyperparameter settings used in these experiments. For all of the above model configurations we use a batch size of 200 during training, except for the scalable NDPPs trained on the Amazon apparel, Amazon three-category, Instacart, and Million Song datasets, where a batch size of 800 is used. |