Value of Information Based on Decision Robustness
Authors: Suming Chen, Arthur Choi, Adnan Darwiche
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We show empirically that the new criterion can reduce the expended budget significantly while reducing the classification accuracy only slightly. We also show empirically that the new criterion leads to decisions that are much more robust than those based on traditional VOI criteria, such as information gain and classification loss. |
| Researcher Affiliation | Academia | Suming Chen, Arthur Choi, and Adnan Darwiche Computer Science Department University of California, Los Angeles {suming,aychoi,darwiche}@cs.ucla.edu |
| Pseudocode | Yes | Algorithm 1: Calc ESDP Computing E-SDP in a Naive Bayes network with class variable D, evidence e, and attributes H = {H1, . . . , Hn}. Algorithm 2: Find Cands Finding candidate subsets to compute E-SDP over. |
| Open Source Code | No | The paper does not provide a link to open-source code or explicitly state that code for their method is available. |
| Open Datasets | Yes | We performed experiments on Naive Bayes networks from a variety of sources: UCI Machine Learning Repository (Bache and Lichman 2013), BFC (http://www.berkeleyfreeclinic.org/) and CRESST (http:// www.cse.ucla.edu/). |
| Dataset Splits | No | The paper does not explicitly provide details about training, validation, and test splits (e.g., percentages or sample counts). It mentions classifying examples in each dataset, but not how the datasets were partitioned for training, validation, or testing. |
| Hardware Specification | No | The paper does not specify the hardware used for running the experiments. |
| Software Dependencies | No | The paper does not list specific software dependencies with version numbers. |
| Experiment Setup | Yes | We experimented with thresholds L ranging from [0.70, 0.75, 0.80, 0.85, 0.90, 0.95]. For each network, the budget was set to 1/3 the number of features, with decision thresholds in [0.1, 0.2, . . . , 0.8, 0.9]. |