Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/iamdecode/lemon
The LEMON machine learning explanation technique 🍋
https://github.com/iamdecode/lemon
Last synced: about 2 months ago
JSON representation
The LEMON machine learning explanation technique 🍋
- Host: GitHub
- URL: https://github.com/iamdecode/lemon
- Owner: iamDecode
- License: bsd-2-clause
- Created: 2019-04-01T14:47:20.000Z (over 5 years ago)
- Default Branch: main
- Last Pushed: 2024-04-10T11:45:35.000Z (9 months ago)
- Last Synced: 2024-11-01T19:32:46.117Z (about 2 months ago)
- Language: Python
- Homepage: https://explaining.ml/lemon
- Size: 69.3 KB
- Stars: 3
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
- Citation: CITATION.cff
Awesome Lists containing this project
README
[](https://badge.fury.io/py/lemon_explainer)
**LEMON** is a technique to explain why predictions of machine learning models are made. It does so by providing feature contribution: a score for each feature that indicates how much it contributed to the final prediction. More precisely, it shows the sensitivity of the feature: a small change in an important feature's value results in a relatively large change in prediction. It is similar to the popular [LIME](https://github.com/marcotcr/lime) explanation technique, but is more faithful to the reference model, especially for larger datasets.
[Website ↗](https://explaining.ml/lemon)
[Academic paper ↗](https://link.springer.com/chapter/10.1007/978-3-031-30047-9_7)
## Installation
To install use pip:
```
$ pip install lemon-explainer
```
## Example
A minimal working example is shown below:
```python
import numpy as np
import pandas as pd
from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
from lemon import LemonExplainer
# Load dataset
data = load_iris(as_frame=True)
X = data.data
y = pd.Series(np.array(data.target_names)[data.target])
# Train complex model
clf = RandomForestClassifier()
clf.fit(X, y)
# Explain instance
explainer = LemonExplainer(X, radius_max=0.5)
instance = X.iloc[-1, :]
explanation = explainer.explain_instance(instance, clf.predict_proba)[0]
explanation.show_in_notebook()
```
## Development
For a development installation (requires npm or yarn),
```
$ git clone https://github.com/iamDecode/lemon.git
$ cd lemon
```
You may want to (create and) activate a virtual environment:
```
$ python3 -m venv venv
$ source venv/bin/activate
```
Install requirements:
```
$ pip install -r requirements.txt
```
And run the tests with:
```
$ pytest .
```
## Approximate distance kernel LIME
If you prefer to use a Gaussian distance kernel as used in LIME, we can approximate this behavior with:
```python
from lemon import LemonExplainer, gaussian_kernel
from scipy.special import gammainccinv
DIMENSIONS = X.shape[1]
KERNEL_SIZE = np.sqrt(DIMENSIONS) * .75 # kernel size as used in LIME
# Obtain a distance kernel very close to LIME's gaussian kernel, see the paper for details.
p = 0.999
radius = KERNEL_SIZE * np.sqrt(2 * gammainccinv(DIMENSIONS / 2, (1 - p)))
kernel = lambda x: gaussian_kernel(x, KERNEL_SIZE)
explainer = LemonExplainer(X, distance_kernel=kernel, radius_max=radius)
```
This behavior is as close as possible to LIME, but still yields more faithful explanations due to LEMON's improved sampling technique. Read the paper for more details about this approach.
## Citation
If you want to refer to our explanation technique, please cite our paper using the following BibTeX entry:
```bibtex
@inproceedings{collaris2023lemon,
title={{LEMON}: Alternative Sampling for More Faithful Explanation Through Local Surrogate Models},
author={Collaris, Dennis and Gajane, Pratik and Jorritsma, Joost and van Wijk, Jarke J and Pechenizkiy, Mykola},
booktitle={Advances in Intelligent Data Analysis XXI: 21st International Symposium on Intelligent Data Analysis (IDA 2023)},
pages={77--90},
year={2023},
organization={Springer}
}
```
## License
This project is licensed under the BSD 2-Clause License - see the [LICENSE](LICENSE) file for details.