Utrecht University

This course focuses on the applications of machine learning algorithms to real-world questions. The overall aim is to provide theories, techniques, tools, and practical experience for applying machine learning to tackle data science problems. The course lectures cover five parts: essential concepts and techniques of machine learning, classification, regression, and clustering; application - outlier detection; application - predictive process mining; application - natural language processing; and application - reinforcement learning. For each of the four application areas, students work in a team to conduct an assignment that applies machine learning algorithms to a real-world dataset.

This course studies utopian and speculative literature as narrative tools to imagine the future. Students learn that these utopian texts reflect a historical setting and mind set. The course studies the function and meaning of utopian texts at two turning points in history: the age of colonialism and the scientific revolution (sixteenth through eighteenth century) and the social-economic tensions and changes in the late nineteenth and early twentieth centuries. Central in these two periods is the focus on the interplay between the European and non-European visions on possible futures. In the early modern period, utopian writers and thinkers have to adapt to a broader geographical (The New World) and philosophical (a New World view) perspective. They have to deal with their role as colonizers (cultural superiority vs. cultural relativism) and scientists (positivism vs. skepticism). In the second period, utopian writing itself is becoming a global endeavor, and often takes the shape of a literary dialogue between former colonizing and colonized countries. In both periods the role of utopias and dystopias in social and political constellations is addressed. Students consider how literature intervenes in conflicts and debates on science, religion, and politics; how utopian optimism or irony can develop into pessimism and (dystopian) skepticism. 

In this course, we build on foundation from database systems, focusing on two important issues. The first issue concerns how to deal with large volumes of data that do not have the precise record structure found in databases. The amount of unstructured data (primarily text) in the world far exceeds the amount of structured data. Searching through text requires a very different approach, especially because the number of results can be extremely large, making ranking based on relevance essential. This field is known as Information Retrieval (IR). Although this discipline has existed for quite some time, its relevance has increased in recent years due to the demand for web search engines. Become familiar with basic IR concepts such as precision, recall, Boolean search, indexing and posting lists, term weighting, the vector space model, and relevance feedback. Also take a detailed look at Google’s PageRank algorithm. This part includes a practical assignment in which IR techniques are applied to processing queries on relational databases, addressing the problem that the number of results can be either too large or too small. the second issue is how to extract interesting patterns and models from data. This is the domain of data mining and machine learning. Here too, the emphasis is on the analysis of unstructured data (again, primarily text), such as using data mining for document classification and clustering, as well as for ranking documents based on their relevance to a given query. The term “document” should be interpreted broadly: it may refer to web pages, email messages (spam or not spam?), posts to a newsgroup, or even tweets. The techniques covered include, among others, Naive Bayes classification, nearest neighbor, support vector machines, hierarchical clustering, and partitioning methods such as k-means clustering. This part also includes a practical assignment in which the data analysis techniques discussed in the lectures be applied to problems as described above. For this, we use the data analysis system R. Assumed previous knowledge in Databases (INFODB), Graphics (INFOGR), and Research Methods in Computer Science or Game Technology. If you have not passed these courses (or other courses in which you acquired comparable prior knowledge), we advise you not to choose this course.

This course examines the basic knowledge and skills needed to give an academic presentation and to interact with others in an English-language classroom. It also covers how to modify the pronunciation of English in order to be better understood by both native and non-native speakers of English; and how to recognize and understand a number of well-known native and non-native accents of English.