Events

The objective of my talk is to introduce the audience to the exciting field of computational fabrication. The recent, wide availability of 3D printers has triggered considerable interest in academia and industry. Computer scientists could engage with hands-on 3D printing and very soon realize the immense but untapped potential of the manufacturing industry for computational methods. In this talk, I will explain the principles of 3D printing (also known as additive manufacturing) both from hardware and software viewpoints. I will then show examples of recent research addressing computational problems in both 3D printing and general manufacturing. I will also discuss my main research interest, i.e. computational fabrication of visual appearance. Appearance of objects is among their most important and most complicated properties that influence or in numerous cases define their function. I show that additive manufacturing provides unprecedented opportunities to create products with novel and useful appearance properties.

Amortized analysis is a standard algorithmic technique for estimating upper bounds on the average costs of functions, specifically operations on data structures. This thesis intends to develop λ-amor, a type-theory for amortized analysis of higher-order functional programs. A typical amortized analysis works by storing ghost resource called /potential/ with a data structure's internal state. Accordingly, the central idea in λ-amor is a type-theoretic construct to associate potential with an arbitrary type. Additionally, λ-amor relies on standard concepts from substructural and modal type systems: indexed monads, affine types and indexed exponential types. We show that λ-amor is not only sound (in a very elementary logical relations model), but also very expressive: It can be used to analyze both eager and lazy data structures, and it can embed existing resource analysis frameworks. In fact, λ-amor is /complete/ for the cost analysis of lazy PCF programs. Further, the basic principles behind λ-amor can be adapted (by dropping affineness and adding mutable state) to obtain an expressive type system for a completely unrelated application, namely, information flow control.

The proposal talk will cover the broad setting and the motivation of the work and a significant subset of λ-amor, but due to time constraints, it will not cover all of λ-amor or the adaptation to information flow control. Implementation of the two type theories is not in the scope of the thesis.

Learning technologies are becoming increasingly important in today's education. This includes game-based learning and simulations, which produce high volume output, and MOOCs (massive open online courses), which reach a broad and diverse audience at scale. The users of such systems often are of very different backgrounds, for example in terms of age, prior knowledge, and learning speed. Adaptation to the specific needs of the individual user is therefore essential. In this talk, I will present two of my contributions on modeling and predicting student learning in computer-based environments with the goal to enable individualization. The first contribution introduces a new model and algorithm for representing and predicting student knowledge. The new approach is efficient and has been demonstrated to outperform previous work regarding prediction accuracy. The second contribution introduces models, which are able to not only take into account the accuracy of the user, but also the inquiry strategies of the user, improving prediction of future learning. Furthermore, students can be clustered into groups with different strategies and targeted interventions can be designed based on these strategies. Finally, I will also describe lines of future research.