Context-based Unsupervised Data Fusion for Decision Making
Authors: Erfan Soltanmohammadi, Mort Naraghi-Pour, Mihaela Schaar
ICML 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Numerical results show the dramatic improvement of the proposed method as compared with the state of the art approaches. In this section, first we use a system with up to 8 classifiers to evaluate the performance of the proposed approach. |
| Researcher Affiliation | Collaboration | Erfan Soltanmohammadi ERFAN@MARVELL.COM Marvell Semiconductor, Inc., Santa Clara, CA, USA. Mort Naraghi-Pour NARAGHI@LSU.EDU Louisiana State University, Baton Rouge, LA, USA. Mihaela van der Schaar MIHAELA@EE.UCLA.EDU University of California, Los Angeles, CA, USA. |
| Pseudocode | Yes | Algorithm 1 Estimation of the parameter set and FC s decisions |
| Open Source Code | No | No statement about providing open-source code or a link to a code repository is found in the paper. |
| Open Datasets | Yes | In order to evaluate the performance of the proposed approach for real data, we used the Wisconsin breast cancer data set (Murphy & Aha, 1994). |
| Dataset Splits | No | The paper mentions "training and validation" in the context of other methods (e.g., "requires large training and validation data sets (Huang & Suen, 1995)") but does not specify train/validation/test splits for its own experimental setup for reproducibility. |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory) used for running experiments are mentioned in the paper. |
| Software Dependencies | No | The paper mentions using "machine learning classifiers from Weka" and cites a book for detailed descriptions, but it does not provide specific version numbers for Weka or any other software dependencies required to replicate the experiment. |
| Experiment Setup | Yes | The parameter T is chosen to be 100. We initialize the EM algorithm with all the probabilities of false alarm equal to 0.2, all the probabilities of detection equal to 0.8, and φ1(t) = 0.6 for t = t0, t0 + 1, , T + t0 1. Here we set cik = c for all i = 0, 1 and k = 1, 2, , K. Three different values of c = 0.2, 1.7, 3.2 are used. for each i and k we set the Lipschitz constants cik = 0.058. |