AutoML - Complete Machine Learning Library

A comprehensive ML library that automates the entire pipeline: Data → EDA → Preprocessing → Split → Model → Evaluate → Predict

Installation

pip install automl-quickpipe

Quick Start

import pandas as pd
from automl import AutoML

# Load your dataset
df = pd.read_csv('your_data.csv')

# Initialize and run complete analysis
ml = AutoML(df, target_col='target_column')
ml.run_complete_analysis()

# Make predictions
new_data = [features]
ml.predict_new(new_data)

Features

• Automated EDA: Statistical summaries, missing value detection, correlation matrices, and distribution plots.
• Smart Preprocessing: Automatic handling of missing values, label encoding for categorical data, and feature scaling.
• Problem Detection: Automatically detects whether to use Classification or Regression based on the target column.
• Multiple Algorithms: Trains and compares multiple models (KNN, Decision Trees, Random Forest, SVM, etc.).
• Model Evaluation: Comprehensive metrics (Accuracy, F1-Score, R², RMSE) with cross-validation.
• Best Model Selection: Automatically identifies and uses the best-performing model for predictions.

Requirements

• numpy
• pandas
• matplotlib
• seaborn
• scikit-learn
• ipython (for Jupyter notebook support)

License

MIT