https://github.com/tsudalab/dt-sampler

https://github.com/tsudalab/dt-sampler

Last synced: about 1 month ago
JSON representation

• Host: GitHub
• URL: https://github.com/tsudalab/dt-sampler
• Owner: tsudalab
• Created: 2023-07-17T05:37:51.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-09-02T17:42:02.000Z (over 1 year ago)
• Last Synced: 2023-09-03T11:17:41.608Z (over 1 year ago)
• Language: Python
• Size: 144 KB
• Stars: 1
• Watchers: 8
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# DT-Sampler

  

## Details

You can find more details about DT-Sampler at [https://arxiv.org/abs/2307.13333](https://arxiv.org/abs/2307.13333).

  

## Abstract

DT-sampler is an ensemble model based on decision tree sampling. Different from random forest, DT-sampler uniformly samples decision trees from a given space, which can generate more stable results and provide higher interpretability compared to random forest. DT-sampler only has two key parameters: #node and threshold. #node constrains the size of decision trees generated by DT-sampler and threshold ensures a minimum training accuracy for each decision tree.







① Encode the construction of decision trees as a SAT problem. \

② Utilize SAT sampler to uniformly sample multiple satisfiable solutions from the high accuracy space.\

③ Decode the satisfiable solutions back into decision trees.\

④ Estimate the training accuracy distribution of the decision trees in the high accuracy space.\

⑤ Measure feature importance by calculating the emergence probability  of each feature.

## Requirements

matplotlib == 3.6.3 \

numpy == 1.21.0 \

pandas == 1.5.3 \

pyunigen == 2.5.2 \

scikit_learn == 1.2.1 \

scipy == 1.11.1 \

z3_solver == 4.12.1.0

## Quick Start

```python

...

dt_sampler = DT_sampler(X_train, y_train, #node, threshod, "./cnf/cnf_name.cnf")

dt_sampler.run(#tree, method = "unigen", seed)

...

```

## Contact

Chao Huang ([email protected])\

Department of Computational Biology and Medical Science\

The University of Tokyo