Language Models

Implemented Markov (n-grams) and perceptrons (neural network) based models for text prediction, recognition, and classification.

The Markov model-based program was able to predict the next word based on the history of the preceding words resulting in context-sensitive word suggestions for better user experience in text-based applications.

The perception-based model imitates the human brain processing information. It was more useful for accurate text recognition from images (I fed it with the scans of digits) achieving high accuracy in categorizing unstructured textual data.