Topic
Natural Language Processing
Exploring how machine learning models represent, classify, and learn patterns from human language.
Natural Language Processing explores how machines can transform human language into numerical representations that can be processed by learning algorithms.
My initial exploration focused on text classification and sentiment analysis, comparing traditional machine-learning representations with neural approaches.
A central question was how text—an inherently symbolic and variable-length form of information—can be converted into representations that preserve enough meaning for a model to make useful predictions.
This led me through techniques including text preprocessing, bag-of-words-style representations, learned embeddings, pretrained word vectors, recurrent networks, and bidirectional sequence models.
This topic established the foundation for my later exploration of tokenization, Transformers, and large language models.