Inhalt

[ 993MLPELRNV19 ] VL LSTM and Recurrent Neural Nets

Versionsauswahl
Workload Education level Study areas Responsible person Hours per week Coordinating university
3 ECTS M1 - Master's programme 1. year (*)Artificial Intelligence Sepp Hochreiter 2 hpw Johannes Kepler University Linz
Detailed information
Original study plan Master's programme Artificial Intelligence 2021W
Objectives In this class, students will learn about Recurrent neural networks (RNNs) and Long Short Term Memory (LSTM), types of neural networks that are very commonly used for predicting sequential data, such as time-series, text or DNA. We will cover both theoretical fundamentals and advanced applications. It is expected that students visiting this class already have a solid understanding of deep learning.
Subject
  • Basics of sequence prediction
  • Recurrent neural networks
  • Vanishing Gradients and LSTMs
  • Attention Models
  • Memory Models (e.g. Neural Turing Machine)
  • Autoregressive Sequence Models and Transformers (PixelRNN, ByteNet, Transformers)
  • Applications in NLP (e.g. next character prediction, language modelling, sentiment analysis)
Criteria for evaluation Exam
Language English
Study material Lecture notes and slides
Changing subject? No
On-site course
Maximum number of participants -
Assignment procedure Direct assignment