Classify-with-xgboost

How to classify with the XGBoost library using a synthetic data generator.

Overview

This project demonstrates binary and multi-class classification using XGBoost, with synthetic datasets produced by a reusable DataGenerator class backed by scikit-learn.

Files

File	Description
data_generator.py	DataGenerator class – creates synthetic train/test splits
classify.py	Trains an XGBoost classifier and prints accuracy & classification report
requirements.txt	Python dependencies

Installation

pip install -r requirements.txt

Usage

python classify.py

Example output

=== Binary Classification ===
Dataset generated: 1000 samples, 20 features, 2 classes.
Train size: 800, Test size: 200

Accuracy: 0.9350

Classification Report:
              precision    recall  f1-score   support
           0       0.95      0.92      0.93        95
           1       0.93      0.95      0.94       105
    accuracy                           0.94       200

=== Multi-Class Classification (3 classes) ===
Dataset generated: 1000 samples, 20 features, 3 classes.
Train size: 800, Test size: 200

Accuracy: 0.7950
...

Using DataGenerator in your own code

from data_generator import DataGenerator
import xgboost as xgb

gen = DataGenerator(n_samples=2000, n_features=15, n_informative=8, n_classes=2)
X_train, X_test, y_train, y_test = gen.generate()

model = xgb.XGBClassifier(n_estimators=100, max_depth=6, learning_rate=0.1)
model.fit(X_train, y_train)
print("Accuracy:", model.score(X_test, y_test))