| Name | vaganboost JSON |
| Version |
0.7.7
JSON |
| download |
| home_page | https://github.com/AliBavarchee/vaganboost |
| Summary | A hybrid model combining VAE, GAN, and LightGBM for boosting performance in high-energy physics or data analysis tasks. |
| upload_time | 2025-02-02 21:03:17 |
| maintainer | None |
| docs_url | None |
| author | Ali Bavarchee |
| requires_python | >=3.6 |
| license | None |
| keywords |
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# VaganBoost: Hybrid VAE-GAN + LightGBM for Advanced Classification 0.7.6
## Introduction
VAGANBoost is a hybrid generative model combining Variational Autoencoders (VAE) and Generative Adversarial Networks (GAN) with boosting techniques to enhance high-energy gamma-ray analysis.
## Outlines
- Implements cVAE+cGAN and cGAN+cVAE+RandomForest models
- Designed for high-energy physics applications
- Utilizes deep learning and gradient boosting techniques
## Key Features
- **Hybrid Architecture**: Combines deep generative models with gradient boosting
- **VAE-GAN Integration**: Joint latent space learning for improved feature representation
- **LightGBM Classifier**: State-of-the-art gradient boosting for final classification
- **Automatic Feature Fusion**: Combines VAE latent features with GAN-generated features
- **Visualization Tools**: Built-in metrics visualization and feature analysis
- **PyTorch Backend**: GPU-accelerated training with seamless CUDA support
## Key Features Table
| Feature | Description | Benefit |
|---------|-------------|---------|
| **VAE-GAN Fusion** | Combines reconstruction power of VAEs with GANs' generative capabilities | Enhanced feature learning |
| **LightGBM Integration** | Gradient boosting on learned features | Superior classification performance |
| **Automatic GPU Support** | Seamless CUDA integration | Faster training on supported hardware |
| **Dynamic Feature Fusion** | Combines latent and generated features | Improved representation learning |
| **Visualization Suite** | Built-in metrics plotting | Easy model evaluation |
## Troubleshooting
**Common Issues:**
1. **CUDA Out of Memory**: Reduce batch size or input dimensions
2. **Poor Classification Performance**:
- Increase VAE latent dimensions
- Adjust GAN-LightGBM feature ratio
3. **Training Instability**:
```python
model = VaganBoost(
...,
vae_kl_weight=0.5, # Adjust KL loss weight
gan_gp_weight=10.0 # Add gradient penalty
)
## Installation
### Prerequisites
- Python 3.6+
- NVIDIA GPU (recommended) with CUDA 11.0+
### Install via pip
```bash
pip install vaganboost
```
### From source
```bash
git clone https://github.com/AliBavarchee/vaganboost.git
cd vaganboost
pip install -e .
```
## Quick Start
### Basic Usage
```python
from vaganboost import VaganBoost, load_data, split_data, normalize_data
# Prepare data
X, y = load_data("data.csv", target_column="label")
X_train, X_test, y_train, y_test = split_data(X, y, test_size=0.2)
X_train_norm, X_test_norm = normalize_data(X_train, X_test)
# Initialize model
model = VaganBoost(
vae_input_dim=X_train_norm.shape[1],
vae_latent_dim=64,
gan_input_dim=100,
num_class=4,
device="cuda"
)
# Train components
model.train_vae(X_train_norm, epochs=100)
model.train_gan(X_train_norm, epochs=50)
model.train_lgbm(X_train_norm, y_train)
# Evaluate
accuracy = model.evaluate(X_test_norm, y_test)
print(f"Test Accuracy: {accuracy:.2%}")
```
### Advanced Configuration
```python
# Custom LightGBM parameters
lgbm_params = {
'objective': 'multiclass',
'num_class': 4,
'metric': 'multi_logloss',
'num_leaves': 63,
'learning_rate': 0.1,
'feature_fraction': 0.7
}
model = VaganBoost(
vae_input_dim=128,
vae_latent_dim=64,
gan_input_dim=100,
num_class=4,
lgbm_params=lgbm_params,
device="cuda"
)
```
## Documentation
### Core Components
| Module | Description |
|--------|-------------|
| `data_utils` | Data loading, splitting, and normalization |
| `models` | VAE, GAN, and LightGBM implementations |
| `train` | Joint training procedures |
| `utils` | Visualization and evaluation tools |
## Dependencies
See `requirements.txt` for required packages.
## License
[](https://opensource.org/licenses/MIT)
This project is licensed under the MIT License.
[](https://www.python.org/downloads/)
Contact
Ali Bavarchee - ali.bavarchee@gmail.com
Project Link: https://github.com/AliBavarchee/vaganboost
=============================================<p align="Center"></p>=============================================
=====
----
| https://www.linkedin.com/in/ali-bavarchee-qip/ |
Raw data
{
"_id": null,
"home_page": "https://github.com/AliBavarchee/vaganboost",
"name": "vaganboost",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.6",
"maintainer_email": null,
"keywords": null,
"author": "Ali Bavarchee",
"author_email": "ali.bavarchee@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/9b/5d/f9f8613443c22645c6b839089041e7838940effb1a781033e739c5848b63/vaganboost-0.7.7.tar.gz",
"platform": null,
"description": "# VaganBoost: Hybrid VAE-GAN + LightGBM for Advanced Classification 0.7.6\r\n\r\n\r\n\r\n## Introduction\r\nVAGANBoost is a hybrid generative model combining Variational Autoencoders (VAE) and Generative Adversarial Networks (GAN) with boosting techniques to enhance high-energy gamma-ray analysis.\r\n\r\n## Outlines\r\n- Implements cVAE+cGAN and cGAN+cVAE+RandomForest models\r\n- Designed for high-energy physics applications\r\n- Utilizes deep learning and gradient boosting techniques\r\n\r\n\r\n\r\n\r\n\r\n## Key Features\r\n\r\n- **Hybrid Architecture**: Combines deep generative models with gradient boosting\r\n- **VAE-GAN Integration**: Joint latent space learning for improved feature representation\r\n- **LightGBM Classifier**: State-of-the-art gradient boosting for final classification\r\n- **Automatic Feature Fusion**: Combines VAE latent features with GAN-generated features\r\n- **Visualization Tools**: Built-in metrics visualization and feature analysis\r\n- **PyTorch Backend**: GPU-accelerated training with seamless CUDA support\r\n\r\n\r\n## Key Features Table\r\n\r\n| Feature | Description | Benefit |\r\n|---------|-------------|---------|\r\n| **VAE-GAN Fusion** | Combines reconstruction power of VAEs with GANs' generative capabilities | Enhanced feature learning |\r\n| **LightGBM Integration** | Gradient boosting on learned features | Superior classification performance |\r\n| **Automatic GPU Support** | Seamless CUDA integration | Faster training on supported hardware |\r\n| **Dynamic Feature Fusion** | Combines latent and generated features | Improved representation learning |\r\n| **Visualization Suite** | Built-in metrics plotting | Easy model evaluation |\r\n\r\n## Troubleshooting\r\n\r\n**Common Issues:**\r\n1. **CUDA Out of Memory**: Reduce batch size or input dimensions\r\n2. **Poor Classification Performance**: \r\n - Increase VAE latent dimensions\r\n - Adjust GAN-LightGBM feature ratio\r\n3. **Training Instability**:\r\n ```python\r\n model = VaganBoost(\r\n ...,\r\n vae_kl_weight=0.5, # Adjust KL loss weight\r\n gan_gp_weight=10.0 # Add gradient penalty\r\n )\r\n\r\n## Installation\r\n\r\n### Prerequisites\r\n- Python 3.6+\r\n- NVIDIA GPU (recommended) with CUDA 11.0+\r\n\r\n### Install via pip\r\n```bash\r\npip install vaganboost\r\n```\r\n\r\n### From source\r\n```bash\r\ngit clone https://github.com/AliBavarchee/vaganboost.git\r\ncd vaganboost\r\npip install -e .\r\n```\r\n\r\n## Quick Start\r\n\r\n### Basic Usage\r\n```python\r\nfrom vaganboost import VaganBoost, load_data, split_data, normalize_data\r\n\r\n# Prepare data\r\nX, y = load_data(\"data.csv\", target_column=\"label\")\r\nX_train, X_test, y_train, y_test = split_data(X, y, test_size=0.2)\r\nX_train_norm, X_test_norm = normalize_data(X_train, X_test)\r\n\r\n# Initialize model\r\nmodel = VaganBoost(\r\n vae_input_dim=X_train_norm.shape[1],\r\n vae_latent_dim=64,\r\n gan_input_dim=100,\r\n num_class=4,\r\n device=\"cuda\"\r\n)\r\n\r\n# Train components\r\nmodel.train_vae(X_train_norm, epochs=100)\r\nmodel.train_gan(X_train_norm, epochs=50)\r\nmodel.train_lgbm(X_train_norm, y_train)\r\n\r\n# Evaluate\r\naccuracy = model.evaluate(X_test_norm, y_test)\r\nprint(f\"Test Accuracy: {accuracy:.2%}\")\r\n```\r\n\r\n### Advanced Configuration\r\n```python\r\n# Custom LightGBM parameters\r\nlgbm_params = {\r\n 'objective': 'multiclass',\r\n 'num_class': 4,\r\n 'metric': 'multi_logloss',\r\n 'num_leaves': 63,\r\n 'learning_rate': 0.1,\r\n 'feature_fraction': 0.7\r\n}\r\n\r\nmodel = VaganBoost(\r\n vae_input_dim=128,\r\n vae_latent_dim=64,\r\n gan_input_dim=100,\r\n num_class=4,\r\n lgbm_params=lgbm_params,\r\n device=\"cuda\"\r\n)\r\n```\r\n\r\n## Documentation\r\n\r\n### Core Components\r\n| Module | Description |\r\n|--------|-------------|\r\n| `data_utils` | Data loading, splitting, and normalization |\r\n| `models` | VAE, GAN, and LightGBM implementations |\r\n| `train` | Joint training procedures |\r\n| `utils` | Visualization and evaluation tools |\r\n\r\n## Dependencies\r\nSee `requirements.txt` for required packages.\r\n\r\n## License\r\n[](https://opensource.org/licenses/MIT)\r\n\r\nThis project is licensed under the MIT License.\r\n\r\n[](https://www.python.org/downloads/)\r\n\r\nContact\r\nAli Bavarchee - ali.bavarchee@gmail.com\r\n\r\nProject Link: https://github.com/AliBavarchee/vaganboost\r\n\r\n\r\n=============================================<p align=\"Center\"></p>=============================================\r\n=====\r\n----\r\n| https://www.linkedin.com/in/ali-bavarchee-qip/ |\r\n",
"bugtrack_url": null,
"license": null,
"summary": "A hybrid model combining VAE, GAN, and LightGBM for boosting performance in high-energy physics or data analysis tasks.",
"version": "0.7.7",
"project_urls": {
"Homepage": "https://github.com/AliBavarchee/vaganboost"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "72f09b821fb5093b988ab387fe95310cf3a014c4abdc6d5bbad6c6e8d0fde2ff",
"md5": "09e84a7a7045eb4e28c7ea53f4612fe5",
"sha256": "78362f7fc7798d4d440a13d00b02981c75637782f57accd6995514f243f07939"
},
"downloads": -1,
"filename": "vaganboost-0.7.7-py3-none-any.whl",
"has_sig": false,
"md5_digest": "09e84a7a7045eb4e28c7ea53f4612fe5",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.6",
"size": 16113,
"upload_time": "2025-02-02T21:03:14",
"upload_time_iso_8601": "2025-02-02T21:03:14.919159Z",
"url": "https://files.pythonhosted.org/packages/72/f0/9b821fb5093b988ab387fe95310cf3a014c4abdc6d5bbad6c6e8d0fde2ff/vaganboost-0.7.7-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "9b5df9f8613443c22645c6b839089041e7838940effb1a781033e739c5848b63",
"md5": "92af32254e328231b074650712b0cc7c",
"sha256": "fbdae0167a8f53391a1d3798c3d005280adf7f1a153dcfe300766153e378ca39"
},
"downloads": -1,
"filename": "vaganboost-0.7.7.tar.gz",
"has_sig": false,
"md5_digest": "92af32254e328231b074650712b0cc7c",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.6",
"size": 14540,
"upload_time": "2025-02-02T21:03:17",
"upload_time_iso_8601": "2025-02-02T21:03:17.041307Z",
"url": "https://files.pythonhosted.org/packages/9b/5d/f9f8613443c22645c6b839089041e7838940effb1a781033e739c5848b63/vaganboost-0.7.7.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-02-02 21:03:17",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "AliBavarchee",
"github_project": "vaganboost",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "vaganboost"
}
