This ipython notebook is working to build a model which will predict the gender based on the names.
The dataset used has been taken from here: Kaggle: Gender Based Names
You can follow the analysis on kaggle
- Data Collection:
In this project we have use the labeled dataset i.e name labeled with respective gender from Kaggle.
- Data Preprocessing:
1. Lowercase: convert the name to lowercase
2. Spliting: split each character
3. Padding: pad with empty set to make the length of the names same.
4. Encode: encode each char with respective number i.e. a=1, b=2 and so on till z=26. Blank char " "=0.
5. Embeddings: represent each name as a embedding using above encoding.
- Model Training:
1. Train a Bidirectional LSTM
- Model Testing
1. Test the trained model using some test set
- Deployment
2. Deploy the model and use to make a Webapplication. We have used Stremlit and HuggingFace to Deploy.
Streamlit is used to build a front-end for the Gender prediction app and is deployed on huggingface.co and streamlit.