Events: Data- och informationsteknik events at Chalmers University of TechnologyTue, 13 Nov 2018 17:04:54 +0100 Algorithms for Efficient LDA Topic Reconstruction<p>Room EC, floor 3, EDIT-buidling, Hörsalsvägen 11, Johanneberg</p><p>​Welcome to a seminar with Alessandro Panconesi, Computer Science, Sapienza University of Rome, Director Bertinoro Informatics Center (BiCi).</p><div>​ <br /></div> <div><strong>Abstract:</strong> Informally, topic reconstruction is the problem of automatically recovering the topics of a given corpus of documents. LDA (Latent Dirichlet Allocation) is a famous paradigm that has been proposed to tackle the problem.</div> <div><br />We present a novel approach for LDA (Latent Dirichlet Allocation) topic reconstruction. The main technical idea is to show that the distribution over the documents generated by LDA can be transformed into a distribution for a much simpler generative model in which documents are generated from the same set of topics, but have a much simpler structure: documents are single topic and topics are chosen uniformly at random. </div> <div><br /></div> <div>Furthermore, this reduction is approximation preserving, in the sense that approximate distributions --- the only ones we can hope to compute in practice --- are mapped into approximate distribution in the simplified world. This opens up the possibility of efficiently reconstructing LDA topics in a roundabout way. Given the input corpus, compute an approximation of the document distribution generated by LDA, transform it into an approximate distribution for the single-topic world, and run a reconstruction algorithm in the uniform, single-topic world --- a much simpler task than direct LDA reconstruction. </div> <div><br /></div> <div>We show the viability of the approach by giving very simple algorithms for a generalization of two notable cases that have been studied in the literature, $p$-separability and matrix-like topics.<br /><br />Joint work with: Matteo Almanza, Flavio Chierichetti, Andrea Vattani<br /><br /></div> Mallozzi, Computer Science and Engineering<p>Jupiter 520, conference room,</p><p>​Engineering Trustworthy Self-Adaptive Autonomous Systems</p><br /><div>Autonomous Systems (AS) are becoming ubiquitous in our society. Some examples are autonomous vehicles, unmanned aerial vehicles (UAV), autonomous trading systems, self-managing Telecom networks and smart factories. Autonomous Systems are based on a continuous interaction with the environment in which they are deployed, and more often than not this environment can be dynamic and partially unknown. AS must be able to take decisions autonomously at run-time also in presence of uncertainty. Software is the main enabler of AS and it allows the AS to self-adapt in response to changes in the environment and to evolve, via the deployment of new features.</div> <br />Traditionally, software development techniques are based on a complete description at design time of how the system must behave in different environmental conditions. This is no longer effective since the system has to be able to explore and learn from the environment in which it is operating also after its deployment. Reinforcement learning (RL) algorithms discover policies that can lead AS to achieve their goals in a dynamic and unknown environment. The developer does not specify anymore how the system should act in each possible situation but rather the RL algorithm can achieve an optimal behaviour by trial and error. Once trained, the AS will be capable of taking decisions and performing actions autonomously while still learning from the environment. These systems are becoming increasingly powerful, yet this flexibility comes at a cost: the learned policy does not necessarily guarantee safety or the achievement of the goals.<br /><br /><div>This thesis explores the problem of building trustworthy autonomous systems from different angles. Firstly, we have identified the state of the art and challenges of building autonomous systems, with a particular focus on autonomous vehicles. Then, we have analysed how current approaches of formal verification can provide assurances in a System of Systems scenario. Finally, we have proposed methods that combine formal verification with reinforcement learning agents to address two major challenges: how to trust that an autonomous system will be able to achieve its goals and how to ensure that the behaviour of AS is safe. </div> Technology and the Swedish Quantum Computer<p>Scaniasalen, lecture room,</p><p>​​Welcome to a lunch seminar organised by the Information and Communication Technology Area of Advance.</p><div>​ <br /></div> <div><div>The program is scheduled for appr one hour (12.00-13.00), with an additional hour for further discussion and exchange of experiences. During the seminar we offer a light lunch. Please register <strong>by the latest on Thursday 22nd November.</strong></div> <div> </div> <h4 class="chalmersElement-H4"><a href="" target="_blank">REGISTER HERE &gt;</a></h4> <div> </div> <div><strong>Programme:</strong> </div> <div> </div> <div><strong>Hardware of the Swedish quantum computer</strong></div> <div>Philip Krantz, Department of Microtechnology and Nanoscience</div> <div> </div> <div><strong>Modeling and Software for the Swedish Quantum Computer</strong></div> <div>Göran Johansson, Department of Microtechnology and Nanoscience</div> <div> </div> <div><strong>Joint work in the intersection of quantum computing and cryptography </strong></div> <div>Katerina Mitrokotsa, Department of Computer Science and Engineering</div> <div>Giulia Ferrini, Department of Microtechnology and Nanoscience</div> <div> </div> <div> </div> <div> </div></div> Walulya, Computer Science and Engineering<p>ED, lecture hall,</p><p>On Design and Applications of Practical Concurrent Data Structures</p><br /><div>In recent years, multicore systems have become ubiquitous; processors in ultra-low power embedded systems to supercomputers contain multiple cores. This proliferation of multicore processors is having an enormous impact on how we design and implement software systems. Shared-memory multicore processors are systems on which multiple computation threads can execute concurrently with access to shared system resources. However, the access to shared resources needs to be synchronized; which is generally the cause of significant difficulty with utilizing multicore systems.</div> <p></p> In order to efficiently utilize these multicore processors, we need to design and implement concurrent programming abstractions that programmers at all levels of expertise can trivially use for general-purpose applications development. A common abstraction for synchronized access to shared data is a concurrent data structure. Concurrent data structures are challenging to design and implement due to the requirement to be correct, efficient and practical under various application constraints.<p></p> In this thesis, we propose new techniques for designing efficient concurrent data structures and improvements to existing implementations. We explore design approaches that are easy to implement without concern for the programming language or deployment platform. Additionally, we explore how to utilize concurrent data structures in complex applications, especially those with stringent throughput and latency demands such as data stream processing. <p></p> Tuma, Computer Science and Engineering<p>Jupiter 520, conference room,</p><p>​Efficiency and Correctness in Threat Analysis of Software Systems</p><div><br /></div> Bastys, Computer Science and Engineering<p>ED, lecture hall,</p><p>​A Principled Approach to Securing IoT Apps</p><div>IoT apps are becoming increasingly popular as they allow users to manage their digital lives by connecting otherwise unconnected devices and services: cyberphysical “things” such as smart homes, cars, or fitness armbands, to online services such as Google or Dropbox, to social networks such as Facebook or Twitter. IoT apps rely on end-user programming, such that anyone with an active account on the platform can create and publish apps, with the majority of apps being created by third parties. <br /></div> <div>We demonstrate that the most popular IoT app platforms are susceptible to attacks by malicious app makers and suggest short and longterm countermeasures for securing the apps. For short-term protection we rely on access control and suggest the apps to be classified either as exclusively private or exclusively public, disallowing in this way information from private sources to flow to public sinks. <br /></div> <div>For longterm protection we rely on a principled approach for designing information flow controls. Following these principles we define projected security, a variant of noninterference that captures the attacker’s view of an app, and design two mechanisms for enforcing it. A static enforcement based on a flow-sensitive type system may be used by the platform to statically analyze the apps before being published on the app store. This enforcement covers leaks stemming from both explicit and implicit flows, but is not expressive enough to address timing attacks. Hence we design a second enforcement based on a dynamic monitor that covers the timing channels as well. </div> <br /> Schoepe, Computer Science and Engineering<p>ED, lecture hall,</p><p>​Flexible Information-Flow Control</p>​<span><br /><div>As more and more sensitive data is handled by software, its trustworthiness becomes an increasingly important concern. This thesis presents work on ensuring that information processed by computing systems is not disclosed to third parties without the user's permission; i.e. to prevent unwanted flows of information. While this problem is widely studied, proposed rigorous information-flow control approaches that enforce strong security properties like noninterference have yet to see widespread practical use. Conversely, lightweight techniques such as taint tracking are more prevalent in practice, but lack formal underpinnings, making it unclear what guarantees they provide.</div> This thesis aims to shrink the gap between heavyweight information-flow control approaches that have been proven sound and lightweight practical techniques without formal guarantees such as taint tracking. This thesis attempts to reconcile these areas by (a) providing formal foundations to taint tracking approaches, (b) extending information-flow control techniques to more realistic languages and settings, and (c) exploring security policies and mechanisms that fall in between information-flow control and taint tracking and investigating what trade-offs they incur. <span style="display:inline-block"></span></span> Lenberg, Computer Science and Engineering<p>Svea 219, lecture room,</p><p>​Behavioral Software Engineering – Organizational Change Refocused</p>