MDP-Based Cost Sensitive Classification Using Decision Trees
Authors: Shlomi Maliah, Guy Shani
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We provide experiments over a set of benchmarks (Lomax and Vadera 2013) showing that our MDP approach scales well over these benchmarks. |
| Researcher Affiliation | Academia | Software and Information Systems Engineering Ben Gurion University, Israel shlomima@post.bgu.ac.il, shanigu@bgu.ac.il |
| Pseudocode | No | The paper describes the algorithms and their equations (1-5, 7-12) in narrative text and within equations, but does not present them in a structured pseudocode or algorithm block. |
| Open Source Code | No | The paper mentions implementation languages for various methods, including their own in C#, but does not provide concrete access to its source code or state that it is open-source. |
| Open Datasets | Yes | We experiments with all domains from the recent cost sensitive survey (Lomax and Vadera 2013; 2011)... Dataset statistics are reported in Table 5. |
| Dataset Splits | Yes | We use a standard 80% 20% train-test split. |
| Hardware Specification | Yes | The experiments were run on a Windows 10 machine, i5 CPU, and 8GB RAM. |
| Software Dependencies | No | The paper mentions software like Weka and programming languages (C#, C, Java) but does not provide specific version numbers for these or any dependent libraries. |
| Experiment Setup | Yes | We use a standard 80% 20% train-test split. For the larger datasets Heart, Wine and Hepatitis, we learned decision trees only over subsets of the 10 most costly attributes together with the low cost attributes. |