An attempt to use machine learning to recognize handwritten words and translate them into correct unicode equivalents.
- Research appropriate machine learning algorithms.
- Implement machine learning algorithm on basic set.
- Create neural networks with the ability to expand to n layers with m neurons per layer.
- Apply back propagation to neural networks.
- Train feed forward neural networks to act as logic gates.
- Implement machine learning algorithm on training set.
- Gather images to be used for training set.
- Standardize input characters.
- Generate receptors to be used as neural network input.
- Train neural networks using training set.
- Implement machine learning algorithm on real world set.
- Predict new characters using trained neural networks.
- Profit