| Kommentar |
Description:
The ability to understand and work with data has become a highly demanded major requirement both in the field of academics and business. This course offers the opportunity to dive into the area of data science by theoretically and practically exploring its fundamental principles. Several examples, live demos, and hands-on exercises during class train data-analytic thinking and technical abilities so that participants can utilize the conveyed knowledge and tools to solve a specific empirical problem and/or practical task. The independent solving of such problem and/or task as well as the presentation, interpretation, and critical assessment of the results is expected from the participants in a seminar presentation as well as in the seminar paper.
Contents:
The course will start with examples showing why data-analytic thinking and data-driven decision making are important skills in the world of today. It will be explained what data science is and how it helps to understand natural phenomena, explore patterns, and make predictions using data for the purpose of solving complex tasks and problems. After a short refreshment in basic statistics followed by a tutorial into the R statistical software, the principles of data processing will be introduced by discussing the characteristics of structured and unstructured data and comparing the advantages and disadvantages of various database technologies with respect to different data types. Furthermore, the strengths and weaknesses of big data technology and parallel computing compared to classical storage and computing approaches will be carved out. Along the cross-industry standard process for data mining, various methods and concepts of data understanding, preparation, modeling, and evaluation will be explored. Therefore, it will be shown how to obtain and scrape data from various sources, conduct an exploratory data analysis, and assess data quality. Different ways of data preparation, formatting, and integration will be discussed as well. Moreover, various techniques to generate new features based on the initial feature set and to reduce dimensionality in the data by feature selection and extraction will be elicited. In the last part of the course, a variety of machine learning methods for extracting signals, detecting anomalies, modeling outcome, and making predictions will be presented. It will also be shown how to evaluate, interpret, and communicate the results.
Learning Outcomes:
After successful participation, students will have a fundamental theoretical understanding and a broad practical toolkit for solving data-science related tasks. They will be able to obtain, explore, and preprocess data and apply appropriate methods for modeling outcome, uncovering hidden patterns, and making predictions. Furthermore, skills in presenting and writing empirical papers will be trained.
|
Anmelden