Matrix Completion Under Monotonic Single Index Models

Authors: Ravi Sastry Ganti, Laura Balzano, Rebecca Willett

NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical results on synthetic and real-world datasets demonstrate the competitiveness of the proposed approach.
Researcher Affiliation Academia Ravi Ganti Wisconsin Institutes for Discovery UW-Madison gantimahapat@wisc.edu Laura Balzano Electrical Engineering and Computer Sciences University of Michigan Ann Arbor girasole@umich.edu Rebecca Willett Department of Electrical and Computer Engineering UW-Madison rmwillett@wisc.edu
Pseudocode Yes Algorithm 1 Monotonic Matrix Completion (MMC)
Open Source Code No The paper does not provide an explicit statement or a link for the open-sourcing of its own methodology's code. It mentions using 'a standard off-the-shelf implementation from TFOCS [27]' for a baseline, but not for their proposed method.
Open Datasets Yes We performed experimental comparisons on four real world datasets: paper recommendation, Jester3, ML-100k, Cameraman... ML-100k comes with its own training and testing dataset.
Dataset Splits Yes For the Jester-3 dataset we used 5 randomly chosen ratings for each user for training, 5 randomly chosen rating for validation and the remaining for testing. ML-100k comes with its own training and testing dataset. We used 20% of the training data for validation. For the Cameraman and the paper recommendation datasets 20% of the data was used for training, 20% for validation and the rest for testing.
Hardware Specification No The paper does not provide specific details about the hardware used (e.g., GPU/CPU models, memory) to run the experiments.
Software Dependencies No The paper mentions using 'a standard off-the-shelf implementation from TFOCS [27]' but does not provide specific version numbers for any software dependencies used in their experiments.
Experiment Setup Yes For our synthetic experiments we generated a random 30 20 matrix Z of rank 5 by taking the product of two random Gaussian matrices of size n r, and r m, with n = 30, m = 20, r = 5. The matrix M was generated using the function, g (M i,j) = 1/(1 + exp( c Z i,j)), where c > 0. ... Hence, we set T = 50.