👤 Gender Classification Model Based on Name

A Machine Learning & Deep Learning application that predicts whether a given name is Female or Male using character-level TF-IDF Vectorization and a TensorFlow / Keras Neural Network.

🚀 Note: I build this interactive Streamlit frontend for this project with Antigravity.

---

🌟 Key Features

• ⚡ Real-time Single Name Prediction: Instant gender prediction with confidence scores (%) and probability breakdown.
• 📋 Batch Processing: Predict gender for multiple names simultaneously via multi-line text input or CSV file upload (name column).
• 📥 Export Results: One-click download of batch prediction results as a CSV file (gender_classification_results.csv).
• 🧹 Smart Preprocessing: Automatic text cleaning that normalizes case, removes titles, address codes (r/o, c/o, so, do), numbers, and extra spaces.
• 📊 Interactive Analytics & Insights: Sidebar and tabs providing insights into model architecture, n-grams, and training accuracy.

---

🧠 Model Architecture & Machine Learning Pipeline

1. Preprocessing: Cleans raw names by removing relationship/address keywords (r/o, c/o, urf, so, do, ps), digits, and normalizing whitespace.
2. Feature Extraction: Character-level TF-IDF Vectorizer using ngram_range=(2, 5) with a maximum of 10,000 features.
3. Deep Learning Classifier:
   • Layer 1: Dense (64 units, ReLU activation)
   • Layer 2: Dropout (0.3 rate for regularization)
   • Layer 3: Dense (32 units, ReLU activation)
   • Output Layer: Dense (1 unit, Sigmoid activation)
4. Model Performance: Achieves ~87.5% Accuracy on a test split of 125,000+ names.

---

📁 Repository Structure

├── app.py                            # Streamlit Web Application (Built with Antigravity)
├── gender_classification_model.h5    # Pre-trained Keras Model (.h5 format)
├── gender_model.keras                # Keras Model (.keras format)
├── tfidf_vectorizer.pkl              # Saved TF-IDF Character Vectorizer
├── label_encoder.pkl                 # Saved Label Encoder ('f' -> 0, 'm' -> 1)
├── Names_dataset.csv                 # Dataset containing names and gender labels
├── gender-classification-model.ipynb # Training & Exploratory Data Analysis Notebook
├── requirements.txt                  # Python Dependencies
├── .gitignore                        # Git Ignore file
└── README.md                         # Project Documentation

---

🛠️ Installation & Local Setup

1. Clone the Repository

git clone https://github.com/Pranamchand/Gender-Classification-Model-based-on-Name.git
cd Gender-Classification-Model-based-on-Name

2. Create Virtual Environment & Install Dependencies

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

pip install -r requirements.txt

3. Run the Streamlit App

streamlit run app.py

The web application will open automatically in your browser at http://localhost:8501.