Dr. Pierre Machart

Machine Learning on large-scale biological data. Medical Systems Biology (ZMNH, UKE)

Developing deep learning models for automatic (symbolic) music composition. Tensorflow, Python, on-device (iOS) inference.

Development of mathematical models and algorithms for acoustics and audio processing applications.

Practical sessions and evaluation of the “Introduction to Data Mining” course for the 2nd year master students in "Advanced Computer Science and Applications". Classes, seminars, practical sessions of the “Introduction to Machine Learning“ class (kernel methods) for the 1st year master students in ”Computer Science“. Classes, seminars, practical sessions of the ”Introduction to Data Mining“ class for the 3rd year bachelor students and 1st year master stundents in ”Computer Science“.

Subject: Morphological segmentation and classification of musical streams. Active semi-supervised learning using graphical temporal models (Segmental HMM), with applications to contemporary music and environmental sounds.

Quality assessment, development, integration of an information system in a bank.

Subject: Audio classification automatic system (speech and music). Design and development of a complete solution for audio segmentation, using state-of-the-art methods (SVM) and descriptors, with applications to real audiovisual data (TV shows archives).

Prototyping a virtual reality immersive cube.

Statistical Machine Learning, Numerical Optimisation, Large-Scale setting, Classification, Linear Regression, Dimension Reduction. Design, prototyping (MATLAB, Python) of novel algorithms and study of their theoretical guarantees.

Formal and cognitive sciences (psychoacoustics, perception) and humanities (sociology, musicology) applied to music.

Fundamental and applied physics, signal processing, electronics. Software development, database, networking and system architecture.