Kronecker Determinantal Point Processes
Authors: Zelda E. Mariet, Suvrit Sra
NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In order to validate our learning algorithm, we compared KRK-PICARD to JOINT-PICARD and to the Picard iteration (PICARD) on multiple real and synthetic datasets.2 |
| Researcher Affiliation | Academia | Zelda Mariet Massachusetts Institute of Technology Cambridge, MA 02139 zelda@csail.mit.edu Suvrit Sra Massachusetts Institute of Technology Cambridge, MA 02139 suvrit@mit.edu |
| Pseudocode | Yes | Algorithm 1 KRK-PICARD iteration (...) Algorithm 2 Sampling from a DPP kernel L |
| Open Source Code | No | The paper does not include any explicit statement about releasing source code for the methodology, nor does it provide a link to a code repository. |
| Open Datasets | Yes | We compared KRK-PICARD to PICARD and EM [10] on the baby registry dataset (described indepth in [10]) (...) Finally, to evaluate KRK-PICARD on large matrices of real-world data, we train it on data from the GENES [3] dataset |
| Dataset Splits | No | The paper mentions training data and test sets but does not specify explicit training/validation/test splits with percentages, sample counts, or references to predefined splits for their experiments. For synthetic data, it states "sampling 100 subsets from the true kernel with sizes uniformly distributed between 10 and 190" but not how these were split for training/validation/testing. |
| Hardware Specification | Yes | All experiments were repeated 5 times and averaged, using MATLAB on a Linux Mint system with 16GB of RAM and an i7-4710HQ CPU @ 2.50GHz. |
| Software Dependencies | No | The paper mentions "MATLAB" and "Linux Mint" but does not provide specific version numbers for any software dependencies required to reproduce the experiments. |
| Experiment Setup | Yes | Convergence was determined when the objective change dipped below a threshold δ. As one EM iteration takes longer than one Picard iteration but increases the likelihood more, we set δPIC = δKRK = 10 4 and δEM = 10 5. (...) we set the step-sizes to their largest possible values, i.e. a PIC = 1.3 and a KRK = 1.8. |