Course Contents
Theory: Application-oriented basics of machine learning and related areas statistics (descriptive, explorative, inductive), advanced analytics, data mining, data science and big data; basics of machine learning methods, functions and algorithms; development processes; basics of data science principles and techniques: discussion of business scenarios; collection, review and quality evaluation of data; data preparation, feature engineering; application of methods and use of program systems on the basis of examples; identification and evaluation of possible solutions; model selection, optimization, performance-assessment; essential ideas of model integration in decision-making processes, recommendations for actions, system of systems; examples from current research, e.g. predictive maintenance in aviation and production;
Practical group work: Application of basic features of a software development methodology (e.g. scrum); application of theoretical knowledge on a cooperative development task; practical solution development of an industrial challenge through programming and data evaluation (implementation); documentation and presentation of the results;
Learning Outcomes
On successful completion of this module, students should be able to:
- Assess and evaluate basic developments and possible uses of artificial intelligence (machine learning) in engineering applications (e. g. mechanical engineering)
- Differentiate and explain key concepts and (mathematical) methods of machine learning
- Evaluate selected algorithms and models (e.g. from the diagnostic/prognostic domain) with regard to their performance, robustness and quality from an engineering point of view
- Apply learned competencies in the areas of data acquisition and processing, data-based modelling (diagnosis and prognosis) and prescription
- Structure simple and medium analytical tasks independently by means of standardized processes (CRISP/OSA-CBM), realize them with given data and estimate their economic impact (business value)
Literature
Lecture materials are made available throughout the semester on Moodle.
- Hastie, T.; Tibshirani, R.; Friedman, J.: The Elements of Statistical learning. Springer
- James, G.; Witten, D.; Hastie, T.;Tibshirani, R.: An Introduction To Statistical Learning. Springer
- Ertel, W.: Grundkurs künstliche Intelligenz. Springer
- Mitchell T.: Machine Learning. McGraw Hill
- Bishop, C.M.: Recognition and Machine Learning. Springer
- Witten, I.: Data Mining. Elsevier
Preconditions
Programming knowledge in Matlab and/or Python
Expected Number of Participants
approx. 150
Further Grading Information
Assessment methods: 50 % written exam (60 min) and 50 % documentation, program code and oral exam (presentation of results) of a cooperative development task (" Data Quest")
Official Course Description
Research on artificial intelligence (AI) is currently experiencing a considerable drive due to its social and economic significance. The Department of Computer Science at TU Darmstadt in particular occupies one of the leading positions in this field, both nationally and internationally: Algorithms for automatic image analysis for complex traffic situations, intelligent search and rescue robots, development of learning methods for controlling robots and machines for real-time interactions, and development of machine learning methods for use in agriculture to address the issue of world nutrition represent just a few outstanding examples. Research in these topics is complemented by a basic research-oriented teaching program. However, there is a lack of application-oriented teaching for engineers. Due to the immense interdisciplinarity of this key topic and digitization strategies such as Industry 4.0 and Industrial Internet of Things (IIoT) in engineering, our budding engineers of tomorrow should also definitely dedicate themselves to this topic.
The goal of the course is to interactively teach students the theoretical basics and practical skills in Machine Learning and related areas of Advanced Analytics, Data Science and Software Development. In addition to the teaching of theory with exercises and demonstrations, a so-called hackathon as group work will be a part of the overall examination. This is a challenge to accomplish a cooperative software development using the Scrum approach model, where students are given a programming and data analysis task provided by an industry partner. This enables students to creatively approach and solve the basic principles of software development with practical challenges, to establish contact with a potential employer and to discuss the results achieved with the industry partner at a final event.
Additional Information
The lecture is held under participation of Prof. Dr.-Ing. J. Metternich and Prof. Dr.-Ing. M. Weigold.
Online Offerings
moodle
- Lecturer: Uwe Klingauf