Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/networks-learning/human-aligned-calibration
Code for "Human-Aligned Calibration for AI-Assisted Decision Making", NeurIPS 2023
https://github.com/networks-learning/human-aligned-calibration
alignment calibration decision-making decision-making-under-uncertainty explainable-ai explainable-ml human-ai human-ai-teaming machine-learning
Last synced: about 6 hours ago
JSON representation
Code for "Human-Aligned Calibration for AI-Assisted Decision Making", NeurIPS 2023
- Host: GitHub
- URL: https://github.com/networks-learning/human-aligned-calibration
- Owner: Networks-Learning
- License: mit
- Created: 2023-05-22T15:08:14.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-07-19T16:49:18.000Z (over 1 year ago)
- Last Synced: 2024-04-16T02:14:55.123Z (7 months ago)
- Topics: alignment, calibration, decision-making, decision-making-under-uncertainty, explainable-ai, explainable-ml, human-ai, human-ai-teaming, machine-learning
- Language: Python
- Homepage:
- Size: 12.7 KB
- Stars: 3
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# human-aligned-calibration
## Install and download prerequisites
Install required packages with `pip install -r requirements.txt`
Download Human-AI Interactions Dataset and dependency `haiid.py` (used to load dataset in `hac.py`) from https://github.com/kailas-v/human-ai-interactions into the directory `./human_ai_interactions_data` :
```{r, engine='bash'}
├── human_ai_interactions_data
│ ├── haiid.py
│ ├── haiid_dataset.csv
```## Repository Structure
- `hac.py`: code to pre-process the dataset and run experiment on "Art", "Sarcasm", "Cities" and "Census" tasks
- `./plots`: contains sub-directories `./barplot`, `./roc` and `./hist` where corresponding plots are saved to## Running the Experiment
- Run:
```{r, engine='bash'}
python3 hac.py
```
- Output:
- Table of metrics for all tasks printed to console
- Generated plots saved under `./plots` sub-directories