88888888ba                                            88                     88           
88      "8b                                     ,d    88                     88           
88      ,8P                                     88    88                     88           
88aaaaaa8P'  ,adPPYba,   ,adPPYba,  ,adPPYba, MM88MMM 88          ,adPPYYba, 88,dPPYba,   
88""""""8b, a8"     "8a a8"     "8a I8[    ""   88    88          ""     `Y8 88P'    "8a  
88      `8b 8b       d8 8b       d8  `"Y8ba,    88    88          ,adPPPPP88 88       d8  
88      a8P "8a,   ,a8" "8a,   ,a8" aa    ]8I   88,   88          88,    ,88 88b,   ,a8"  
88888888P"   `"YbbdP"'   `"YbbdP"'  `"YbbdP"'   "Y888 88888888888 `"8bbdP"Y8 8Y"Ybbd8"'   

Supervised Learning with Gradient Boosting

Features

Dataset

• Management: load, save, import and information
• Visualization: raw values, heatmap, statistics, PCA
• Filtering: features and class selection, resampling
• Balancing: parametric oversample and undersample

Model

• Training: early stopping support and result plots
• Tuning: automatic hyperparameter optimization
• Inspection: tree graph and feature importance
• Export: generate executable C++/Python code

Evaluation

• Performance: classification report and confusion matrix
• Graphs: metric curves and probability histogram
• Optimize: optimal thresholds with various metrics
• Export: model predictions and output probabilities

Installation

1. Install latest Miniconda from official site
2. Create a new virtual environment:
   1. Install required packages (choose one of the following options):
      • CPU only: conda env create -f environment_cpu.yml
      • GPU acceleration: conda env create -f environment_gpu.yml
   2. Activate the environment: conda activate bl

Usage

Run main script from inside bl environment:

python boostlab.py

Note: If images inside icons folder are modified, runtime resources need to be updated before running the application:

pyside6-rcc resources.qrc -o resources.py