AutoML

A comprehensive Streamlit-based Automated Machine Learning (AutoML) platform for classification tasks. Upload your data and let the system guide you through exploration, preprocessing, model training, and report generation—all with an interactive, code-free UI.

Live Demo

---

Screenshots

Data Upload & Overview

Exploratory Data Analysis

Data Preprocessing

Model Training

Results & Reports

AI-Powered Data Insights

---

Features

1. Data Upload & Overview

• Upload CSV files or use the built-in sample dataset.
• Automatic data profiling: shape, data types, missing values, and class distribution.
• Select your target column for classification.

2. Exploratory Data Analysis (EDA)

• Missing Values: Visualize and analyze columns with missing data.
• Outlier Detection: Identify outliers using IQR or Z-score methods.
• Correlation Analysis: Interactive correlation heatmaps.
• Distributions: Histograms for numerical features, bar charts for categorical features.

3. Data Preprocessing

• Automated Issue Detection: Detects missing values, outliers, high cardinality, class imbalance, and constant features.
• Interactive Fixes: Choose how to handle each issue (impute, drop, cap, encode, etc.).
• Scaling: StandardScaler or MinMaxScaler.
• Encoding: One-Hot or Ordinal encoding for categorical variables.
• Train/Test Split: Configurable split ratio and random seed.

4. Model Training & Hyperparameter Tuning

• Multiple Algorithms: Logistic Regression, Random Forest, Gradient Boosting, SVM, KNN, Decision Tree, AdaBoost, and a Baseline model.
• Hyperparameter Optimization: Grid Search or Randomized Search with cross-validation.
• Class Imbalance Handling: Optional class weights.
• Performance Metrics: Accuracy, Precision, Recall, F1-Score, ROC-AUC, Confusion Matrix, and ROC Curves.

5. Results & Reports

• Model Comparison: Side-by-side metrics and visualizations.
• Best Model Selection: Automatic recommendation based on F1-Score.
• Report Generation: Download comprehensive PDF or HTML reports.

6. AI-Powered Data Insights Chat

• Conversational Analysis: Ask questions about your data in natural language.
• Auto-Generated Visualizations: Request charts and the AI will generate Plotly code.
• Powered by GROQ API: Uses the LLaMA 4 model for intelligent responses.

---

Project Structure

AutoML/
├── app.py                 # Main Streamlit application
├── requirements.txt       # Python dependencies
├── .env                   # Environment variables (GROQ_API_KEY)
├── sample_data/           # Sample dataset for testing
│   └── loan_classification.csv
└── src/
    ├── __init__.py
    ├── chat.py            # AI chat assistant module (GROQ API)
    ├── data_loader.py     # CSV loading utilities
    ├── eda.py             # Exploratory data analysis functions
    ├── models.py          # ML models and training logic
    ├── preprocessing.py   # Data cleaning and transformation
    └── report.py          # PDF/HTML report generation

---

Installation

Prerequisites

• Python 3.8 or higher
• (Optional) GROQ API key for the AI chat feature

Steps

1. Clone the repository:

   git clone <repository-url>
   cd AutoML

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. (Optional) Set up GROQ API key for AI Chat: Create a .env file in the project root:

   GROQ_API_KEY=your_api_key_here
   

---

Usage

1. Run the application:

   streamlit run app.py

2. Open your browser and navigate to http://localhost:8501.

3. Follow the workflow:

   • Upload & Data Info: Load your CSV or use the sample dataset.
   • Exploratory Analysis: Review EDA visualizations.
   • Data Preprocessing: Apply fixes and transformations.
   • Model Training: Select and train models.
   • Results & Reports: Compare models and download reports.
   • Data Insights Chat: Ask questions about your data (requires GROQ API key).

---

Dependencies

Package	Purpose
streamlit	Web application framework
pandas	Data manipulation
numpy	Numerical operations
scikit-learn	Machine learning models and preprocessing
matplotlib	Static plotting (used internally)
seaborn	Statistical visualizations
plotly	Interactive charts
fpdf	PDF report generation
groq	GROQ API client for AI chat
python-dotenv	Environment variable management

---

Configuration

Environment Variables

Variable	Description	Required
GROQ_API_KEY	API key for GROQ (AI Chat feature)	Optional

---

License

This project is licensed under the MIT License.

---