base on Official implementation of MetaTree: Learning a Decision Tree Algorithm with Transformers <h1 align="center"> 🌲 MetaTree 🌲 </h1>
<p align="center"> <b>Learning a Decision Tree Algorithm with Transformers</b> (<a href="https://arxiv.org/abs/2402.03774">Zhuang et al., TMLR 2024</a>).
</p>
<p align="center">
<img src="https://img.shields.io/badge/license-mit-blue.svg">
<img src="https://img.shields.io/badge/python-3.7+-blue">
<img src="https://img.shields.io/pypi/v/metatreelib?color=green">
</p>
<p align="center"> MetaTree is a transformer-based decision tree algorithm. It learns from classical decision tree algorithms (greedy algorithm CART, optimal algorithm GOSDT), for better generalization capabilities.
</p>
## Quickstart -- use MetaTree to generate decision tree models
Model is available at https://huggingface.co/yzhuang/MetaTree
1. Install `metatreelib`:
```bash
pip install metatreelib
# Alternatively,
# clone then pip install -e .
# pip install git+https://github.com/EvanZhuang/MetaTree
```
2. Use MetaTree on your datasets to generate a decision tree model
```python
from metatree.model_metatree import LlamaForMetaTree as MetaTree
from metatree.decision_tree_class import DecisionTree, DecisionTreeForest
from metatree.run_train import preprocess_dimension_patch
from transformers import AutoConfig
import imodels # pip install imodels
# Initialize Model
model_name_or_path = "yzhuang/MetaTree"
config = AutoConfig.from_pretrained(model_name_or_path)
model = MetaTree.from_pretrained(
model_name_or_path,
config=config,
)
decision_tree_forest = DecisionTreeForest()
# Load Datasets
X, y, feature_names = imodels.get_clean_dataset('fico', data_source='imodels')
print("Dataset Shapes X={}, y={}, Num of Classes={}".format(X.shape, y.shape, len(set(y))))
train_idx, test_idx = sklearn.model_selection.train_test_split(range(X.shape[0]), test_size=0.3, random_state=seed)
# Dimension Subsampling
feature_idx = np.random.choice(X.shape[1], 10, replace=False)
X = X[:, feature_idx]
test_X, test_y = X[test_idx], y[test_idx]
# Sample Train and Test Data
subset_idx = random.sample(train_idx, 256)
train_X, train_y = X[subset_idx], y[subset_idx]
input_x = torch.tensor(train_X, dtype=torch.float32)
input_y = torch.nn.functional.one_hot(torch.tensor(train_y)).float()
batch = {"input_x": input_x, "input_y": input_y, "input_y_clean": input_y}
batch = preprocess_dimension_patch(batch, n_feature=10, n_class=10)
model.depth = 2
outputs = model.generate_decision_tree(batch['input_x'], batch['input_y'], depth=model.depth)
decision_tree_forest.add_tree(DecisionTree(auto_dims=outputs.metatree_dimensions, auto_thresholds=outputs.tentative_splits, input_x=batch['input_x'], input_y=batch['input_y'], depth=model.depth))
print("Decision Tree Features: ", [x.argmax(dim=-1) for x in outputs.metatree_dimensions])
print("Decision Tree Thresholds: ", outputs.tentative_splits)
```
3. Inference with the decision tree model
```python
tree_pred = decision_tree_forest.predict(torch.tensor(test_X, dtype=torch.float32))
accuracy = accuracy_score(test_y, tree_pred.argmax(dim=-1).squeeze(0))
print("MetaTree Test Accuracy: ", accuracy)
```
## Example Usage
We show a complete example of using MetaTree at [notebook](examples/example_usage.ipynb)
## Questions?
If you have any questions related to the code or the paper, feel free to reach out to us at
[email protected].
## Citation
If you find our paper and code useful, please cite us:
```r
@misc{zhuang2024learning,
title={Learning a Decision Tree Algorithm with Transformers},
author={Yufan Zhuang and Liyuan Liu and Chandan Singh and Jingbo Shang and Jianfeng Gao},
year={2024},
eprint={2402.03774},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
", Assign "at most 3 tags" to the expected json: {"id":"7709","tags":[]} "only from the tags list I provide: [{"id":77,"name":"3d"},{"id":89,"name":"agent"},{"id":17,"name":"ai"},{"id":54,"name":"algorithm"},{"id":24,"name":"api"},{"id":44,"name":"authentication"},{"id":3,"name":"aws"},{"id":27,"name":"backend"},{"id":60,"name":"benchmark"},{"id":72,"name":"best-practices"},{"id":39,"name":"bitcoin"},{"id":37,"name":"blockchain"},{"id":1,"name":"blog"},{"id":45,"name":"bundler"},{"id":58,"name":"cache"},{"id":21,"name":"chat"},{"id":49,"name":"cicd"},{"id":4,"name":"cli"},{"id":64,"name":"cloud-native"},{"id":48,"name":"cms"},{"id":61,"name":"compiler"},{"id":68,"name":"containerization"},{"id":92,"name":"crm"},{"id":34,"name":"data"},{"id":47,"name":"database"},{"id":8,"name":"declarative-gui "},{"id":9,"name":"deploy-tool"},{"id":53,"name":"desktop-app"},{"id":6,"name":"dev-exp-lib"},{"id":59,"name":"dev-tool"},{"id":13,"name":"ecommerce"},{"id":26,"name":"editor"},{"id":66,"name":"emulator"},{"id":62,"name":"filesystem"},{"id":80,"name":"finance"},{"id":15,"name":"firmware"},{"id":73,"name":"for-fun"},{"id":2,"name":"framework"},{"id":11,"name":"frontend"},{"id":22,"name":"game"},{"id":81,"name":"game-engine "},{"id":23,"name":"graphql"},{"id":84,"name":"gui"},{"id":91,"name":"http"},{"id":5,"name":"http-client"},{"id":51,"name":"iac"},{"id":30,"name":"ide"},{"id":78,"name":"iot"},{"id":40,"name":"json"},{"id":83,"name":"julian"},{"id":38,"name":"k8s"},{"id":31,"name":"language"},{"id":10,"name":"learning-resource"},{"id":33,"name":"lib"},{"id":41,"name":"linter"},{"id":28,"name":"lms"},{"id":16,"name":"logging"},{"id":76,"name":"low-code"},{"id":90,"name":"message-queue"},{"id":42,"name":"mobile-app"},{"id":18,"name":"monitoring"},{"id":36,"name":"networking"},{"id":7,"name":"node-version"},{"id":55,"name":"nosql"},{"id":57,"name":"observability"},{"id":46,"name":"orm"},{"id":52,"name":"os"},{"id":14,"name":"parser"},{"id":74,"name":"react"},{"id":82,"name":"real-time"},{"id":56,"name":"robot"},{"id":65,"name":"runtime"},{"id":32,"name":"sdk"},{"id":71,"name":"search"},{"id":63,"name":"secrets"},{"id":25,"name":"security"},{"id":85,"name":"server"},{"id":86,"name":"serverless"},{"id":70,"name":"storage"},{"id":75,"name":"system-design"},{"id":79,"name":"terminal"},{"id":29,"name":"testing"},{"id":12,"name":"ui"},{"id":50,"name":"ux"},{"id":88,"name":"video"},{"id":20,"name":"web-app"},{"id":35,"name":"web-server"},{"id":43,"name":"webassembly"},{"id":69,"name":"workflow"},{"id":87,"name":"yaml"}]" returns me the "expected json"