School of Arts & Sciences · Department of Computer Science

Deep Learning for
Natural Language Processing

A course that traces how a machine learns to read — from words and tokens to transformers. Every deck is linked below, free to use.

Instructor
Dr. Chadi Helwe
Contact
chadi.helwe@lau.edu.lb

The course builds one idea at a time: how text becomes tokens, how early language models predict the next word, how machine learning and neural networks learn from data, and how embeddings and transformers reshaped what's possible. Each lecture assumes the one before it — read them in order, or jump to the topic you need.

Slides build on two textbooks Speech and Language Processing 3rd ed., draft — Jurafsky & Martin Natural Language Processing: Neural Networks and Large Language Models — Xiao & Zhu

The curriculum

Read as a sequence — each lecture builds on the one before it.

  1. 01

    Introduction to NLP

    what NLP is, and how the field evolved into deep learning

  2. 02

    Words and Tokens

    morphemes, Unicode, regular expressions, and subword tokenization

  3. 03

    N-gram Language Models

    language modeling with n-grams, and how to evaluate them

  4. 04

    Introduction to Machine Learning

    supervised learning, logistic regression, cross-entropy loss, and gradient descent

  5. 05

    Neural Networks

    multi-layer networks, regularization, optimization, and a PyTorch walkthrough

  6. 06

    Word Representations and Embeddings

    lexical semantics, count-based vectors, cosine similarity, and Word2Vec

  7. 07

    Convolutional & Recurrent Neural Networks

    CNNs and RNNs for text — architectures, transfer learning, LSTMs and GRUs

  8. 08

    Transformers

    attention, transformer blocks, training LLMs, sampling, and prompting

  9. 09

    Transformers: NLP Applications

    BERT, text classification, machine translation, and text generation

Found a mistake or a broken link in the slides — or have a question? Email me — corrections and questions are welcome.