Marius Bock

Marius Bock

PhD student in Ubiquitous Computing and Computer Vision at the University of Siegen.

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

portfolio

publications

talks

teaching

Recent Advances in Machine Learning (Summer 2021)

Graduate course, University of Siegen, 2021

Supervision of a team of students applying Deep Learning on Human Activity Recognition. The students were given a dataset of raw sensor data and were tasked to properly preprocess and apply state-of-the-art Deep Learning (as well as for comparison classical Machine Learning approaches) on top of the processed dataset.

Seminar Data Science (Summer 2021)

Undergraduate & Graduate course, University of Siegen, 2021

Supervision of a team of students competing in the Data Mining Cup 2021. Task of the competition was to implement a book recommender system using both transactional and general information about a set of predefined books. The data was provided by the competition itself.

Seminar Data Science (Winter 2021)

Undergraduate & Graduate course, University of Siegen, 2021

Supervision of a team of students conducting experiments using GANs for style transfer on images. In particular, the group applied CycleGANs in order to transfer the style of prominent painters on different images.

Recent Advances in Machine Learning (Summer 2022)

Graduate course, University of Siegen, 2021

Supervision of a team of students applying Deep Learning on Human Activity Recognition. The students were given a large, unlabeled Human Activity Recogntion dataset and were tasked to undergo the process of labeling and preprocessing raw sensor data until it can be used for training a prediction algorithm. Students were then assigned to reach best performing results laying focus on certain classes which were short in execution.

Seminar Data Science (Summer 2021)

Undergraduate & Graduate course, University of Siegen, 2021

Supervision of a team of students investigating the effect of recurrent layers in Human Activity Recognition. Students were tasked to record their own small dataset and run experiments using a network architecture of choice. Goal was to investigate the effect LSTM layers had on prediction performance.

Deep Learning (Winter 2022)

Graduate course, University of Siegen, 2022

Organization and execution of exercises, around 50 students. The course itself gives an introduction to deep learning, describes common building blocks in the network architectures, introduces optimization algorithms for their training, and discusses strategies that improve the generalization.

Recent Advances in Machine Learning (Summer 2023)

Graduate course, University of Siegen, 2023

Supervision of a team of students applying Deep Learning on Human Activity Recognition. The students were given the newly released WEAR dataset and tasked to adavance the current benchmark using either inertial, vision or a combination of both as input data.