(in Polish) Introduction to deep learning for natural language processing

(in Polish) Basic mathematics (calculus, algebra, and statistics) and basic programming in Python.

The course aims to provide the students with a basic understanding of Neural Networks, so they can build and train their own models for Natural Language Processing using Python and TensorFlow.

We start with basic concepts to understand the perceptron and supervised learning algorithms. Then, we see more sophisticated deep learning architectures (e.g., Recurrent and Convolutional Neural Networks) with NLP tasks such as text classification and sentiment analysis.

The course aims to provide the students with a basic understanding of Neural Networks, so they can build and train models using Python and TensorFlow to solve tasks for Natural Language Processing. Specific topics include:

- Basic mathematics for deep learning (gradient descent, matrices, probability theory)

- Basic Python libraries for deep learning (pandas, NumPy, Matplotlib)

- Introduction to machine learning (linear regression, logistic regression)

- The perceptron (activation function, loss function, backpropagation)

- Metrics to evaluate machine learning algorithms (confusion matrix, F1 score)

- Deep learning Python libraries (TensorFlow and Keras)

- Vector space models (bag of words, n-grams, word embeddings)

- Convolutional neural networks

- Recurrent neural networks

- Other deep learning model architectures (LSTM, GRU, and the Transformer)

- Practical examples for NLP (text classification, sentiment analysis)

Chollet, F. (2018), Deep Learning with Python, Manning.

Ganegedara, T. (2018), Natural Language Processing with TensorFlow, Packt.