One-Pass Algorithms for MAP Inference of Nonsymmetric Determinantal Point Processes

Authors: Aravind Reddy, Ryan A. Rossi, Zhao Song, Anup Rao, Tung Mai, Nedim Lipka, Gang Wu, Eunyee Koh, Nesreen Ahmed

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In Section 7, we evaluate our proposed online NDPP MAP Inference algorithms on several datasets and show that they show that they perform comparably to (or even better than) the offline greedy algorithm while using substantially lower memory.
Researcher Affiliation Collaboration 1Northwestern University 2Adobe Research 3Intel Labs
Pseudocode Yes Algorithm 1 Streaming Partition Greedy MAP Inference for low-rank NDPPs; Algorithm 2 ONLINE-LSS: Online MAP Inference for lowrank NDPPs with Stash.; Algorithm 3 ONLINE-2-NEIGHBOR: Local Search over 2-neighborhoods with Stash for Online NDPP MAP Inference.; Algorithm 4 ONLINE-GREEDY: Online Greedy MAP Inference for NDPPs
Open Source Code No The paper does not provide a specific link or explicit statement about the public availability of their source code for the methodology described.
Open Datasets Yes Amazon Apparel: This dataset consists of 14,970 registries (sets) from the apparel category of the Amazon Baby Registries dataset, which is a public dataset that has been used in prior work on NDPPs (Gartrell et al., 2021; 2019).
Dataset Splits No The paper does not provide specific details on how datasets were split into training, validation, and test sets, or describe a cross-validation setup for their experiments.
Hardware Specification Yes All experiments were performed using a standard desktop computer (Quad-Core Intel Core i7, 16 GB RAM)
Software Dependencies No The paper mentions the use of certain algorithms and concepts (e.g., Python), but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes For all of our results in 7, we set k = 8 and choose α = 1.1.