Model Class Reliance for Random Forests
Authors: Gavin Smith, Roberto Mansilla, James Goulding
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section we demonstrate: the ability of RF-MCR2 to compute \ MCR bounds in a tractable fashion; the convergence of the estimators; the ability to recover true MCR values; and the insights derivable on 2 real world datasets (recidivism and breast cancer). |
| Researcher Affiliation | Academia | Gavin Smith N/LAB University of Nottingham Nottingham, UK Roberto Mansilla N/LAB University of Nottingham Nottingham, UK {first.last}@nottingham.ac.uk James Goulding N/LAB University of Nottingham Nottingham, UK |
| Pseudocode | No | The paper describes the computational steps in text but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | A python version of the implementation is available on github: https://github.com/gavin-s-smith/mcrforest. |
| Open Datasets | Yes | First we consider data sampled from a simple generative network... A dataset of 1000 instances is generated from this network... COMPAS Recidivism Modelling... The COMPAS system provides records from a set of defendants from Broward County, Florida labelled with known recidivism scores [12]... The Breast Cancer Wisconsin dataset [19] is a dataset from the UCI ML Repository... |
| Dataset Splits | No | The paper mentions 'mean squared errors... on a held out test set' but does not specify training, validation, or test split percentages or exact counts for all splits. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the experiments. |
| Software Dependencies | No | The paper mentions a 'python version of the implementation' but does not list specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x). |
| Experiment Setup | No | The paper discusses the method and results but does not provide specific experimental setup details such as hyperparameter values (e.g., learning rate, batch size, epochs) or optimizer settings. |