MSc in Computer Science - Specialization in Data-Intensive Systems

Data Visualization

Course content

Data visualization is the science and practice of creating interactive graphical representations from data to support the data’s exploration and presentation. In data analysis, visualization is the method of choice when the analysis objectives cannot be crisply and formally specified – i.e. when looking for regularities (patterns, trends,…) or irregularities (outliers, anomalies,…) in the data in an “I-know-it-when-I-see-it” manner.

In this course, students will get a broad introduction to the field of Data Visualization. Course topics include core concepts of data visualization, data and task taxonomies, data pre-processing, visual encoding and perception, visualization techniques and layout algorithms, interaction techniques and software design patterns for interactive visualization. Students will learn how to design a visualization that is expressive of the data to be shown, effective for the analysis task to be carried out with it, and appropriate for the technical context (display size, screen vs. print, etc.)

In the course project, students will implement a visualization from start to finish – either for a dataset of their own choice or for one of the provided datasets. They will learn how to clean and reduce their data, how to establish visualization requirements, how to iterate through visualization prototypes, how to realize them in software using state-of-the-art visualization tools and libraries, and how to eventually evaluate them to determine the best visualization solution for a given problem. Using review sessions, participants will not only learn to develop their visualization but also to analyze and critique the visualizations of others and to suggest improvements based on the principles and models introduced in the course.

Description of qualifications

Objectives of the course:

At the end of this course, participants will have gained knowledge of data visualization:

• A method of tightly interlinked graphic design, interaction design, and algorithm design (visualization design).

• A computational process that transforms input data into graphic representations (visualization pipeline).

• A tool for user-driven data analysis from preparation, via exploration and confirmation, to presentation (visual analytics).

Or

Deep Learning for Visual Recognition

Course content

Deep learning has enabled today’s AI systems to drive cars autonomously, beat humans in computer games, and paint whatever you tell them to. The deep learning revolution is upon us, transforming many businesses and changing how we write software.

This course is a deep dive into the details of deep learning with a focus on solving computer vision tasks, like image classification, object detection, image segmentation, face recognition, human pose estimation, and image captioning. Computer vision is a prominent field within deep learning.

During the 14-week course, students will learn to implement, train and debug their neural networks and gain a theoretical understanding of neural network architectures and optimization strategies. The weekly hands-on exercises will involve setting up computer vision problems, like image recognition, and applying learning algorithms and practical engineering tricks for training and fine-tuning the networks.

The mandatory course project allows students to apply what they have learned in class to a problem of their interest (in groups of 1-3 students).

Description of qualifications

Objectives of the course

The participants will after the course have detailed knowledge of deep neural networks and will be able to solve complex visual recognition tasks with deep learning.

Data Mining

Course content

The course introduces fundamental data mining models and algorithms with a focus on unsupervised learning. We study basic properties of numerical, sequential, sets, and graph data, and present methods for efficient access and analysis of such data, including basic indexing techniques and algorithms for disk-based data.
Fundamental data mining methods include frequent patterns, clustering, and outlier detection with applications in web mining, text analysis, time series databases, and others. We consider high dimensional and noisy data. The course presents the versatility of graphs for data analysis. We cover mostly unsupervised and random-walk-based techniques for graphs from a data mining perspective, as well as recent developments in self-supervised learning, spectral graph theory, and dimensionality reduction.

Description of qualifications

Objectives of the course

The participants will after the course have detailed knowledge of data mining techniques, in particular unsupervised methods for numerical, sequential, sets, and graph data, including basic index structures for efficient retrieval and analysis.

Advanced-Data Management and Analysis

Course content

The course covers selected techniques for the management and analysis of data represented in different formats. Examples of such data include a variety of multimedia data represented as multidimensional feature vectors as well as sensor measurements. Key topics include data modeling and data mining as well as indexing techniques and associated query processing techniques. The course will cover recent advances in the area.

Description of qualifications

Objectives of the course:

The participants will after the course have detailed knowledge of the management of multidimensionally represented data, including data mining and query processing, and a basis for understanding current research in data analysis and search. The working method of the course will also train the participants to read, understand and present research papers.