Same-Cluster Querying for Overlapping Clusters
Authors: Wasim Huleihel, Arya Mazumdar, Muriel Medard, Soumyabrata Pal
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Finally, we test our algorithms over both synthetic and real-world data, showing their practicality and effectiveness. |
| Researcher Affiliation | Academia | Wasim Huleihel Department of Electrical Engineering Tel-Aviv University Tel-Aviv, Israel 6997801 wasimh@mail.tau.ac.il Arya Mazumdar College of Information & Computer Sciences University of Massachusetts Amherst Amherst, MA 01003 arya@cs.umass.edu Muriel Médard Electrical Engineering & Computer Science Massachusetts Institute of Technology Cambridge, MA 02139 medard@mit.edu Soumyabrata Pal College of Information & Computer Sciences University of Massachusetts Amherst Amherst, MA 01003 soumyabratap@umass.edu |
| Pseudocode | Yes | Algorithm 1 Findmembership; Algorithm 2 Find Similarity; Algorithm 3 Worst-case quantized responses |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available. |
| Open Datasets | Yes | We focus on real-world data from the popular Movielens dataset (https://grouplens.org/datasets/movielens/) for our experiments. |
| Dataset Splits | No | The paper describes experiments involving random sampling of elements (set S) and conducting queries based on these samples to recover clusters. It does not define traditional train/validation/test splits of the dataset for model training or hyperparameter tuning. The experimental setup involves iterating trials for different sample sizes, but this is distinct from a validation set split. |
| Hardware Specification | No | The paper does not specify any details about the hardware (e.g., CPU, GPU, memory, cloud instances) used for running the experiments. |
| Software Dependencies | No | The paper discusses concepts related to software (e.g., 'low-rank matrix completion'), and mentions an example of a specific solver ('CPLEX 12.4') in the context of related work. However, it does not list any specific software dependencies or their version numbers that were used in the paper's own experiments. |
| Experiment Setup | Yes | We ran Algorithm 3 (...) with different values of |S| (number of movies randomly chosen in the first step) and since the movies are sampled randomly, we ran 50 trials for each value of |S|.; Instead of sampling a fixed number of movies a-priori, we randomly select k = 5 movies and make all pairwise queries restricted to those 5 movies. Then, we check if the 5 × 5 gram matrix (...) is of rank 5 and if yes, then we will use that matrix for further calculations. If not, we sample again until we succeed. |