This course focuses on the practical aspects of selecting and evaluating machine learning systems, emphasizing the importance of proper evaluation to ensure optimal model performance. It covers topics such as model selection, experimentation, data handling, classification metrics, and error analysis through examples and implementation exercises. Designed for individuals with basic knowledge of machine learning, the course assumes familiarity with numerical methods and data structures.
Teachers: Ferrer, Luciana
Access here for more information: URL
The course aims to introduce students to essential concepts related to signal, audio and speech processing, covering both traditional approaches and neural network-based techniques. It will provide participants with initial exposure to the fundamentals of signal processing, including filters and Fourier analysis, and then move into current neural network-based methods, such as representation learning using self-supervised learning. In addition, topics such as acoustics and speech production, machine learning and language models will be explored. Applications of the course will cover areas such as sound event recognition, music genre classification, speaker identification, emotion detection and speech recognition.
Teachers: Riera, Pablo
Access here for more information: URL