קולוקוויום וסמינרים

כדי להצטרף לרשימת תפוצה של קולוקוויום מדעי המחשב, אנא בקר בדף מנויים של הרשימה.


Computer Science events calendar in HTTP ICS format for of Google calendars, and for Outlook.

Academic Calendar at Technion site.

קולוקוויום וסמינרים בקרוב

  • Characterization of Cellular Metabolism throughout the Cell Cycle in Cancer: An Integrated Experimental-Computational Approach

    דובר:
    אויניונג אן, הרצאה סמינריונית לדוקטורט
    תאריך:
    יום שלישי, 4.9.2018, 14:00
    מקום:
    טאוב 601
    מנחה:
    Prof. Tomer Shlomi

    Cellular metabolic demands change throughout the cell cycle. Nevertheless, a characterization of how metabolic fluxes adapt to the changing demands throughout the cell cycle is lacking. The rate of metabolic reactions and pathways in living cells, also referred to as metabolic flux, is not a directly measurable quantity. The most direct approach for quantifying intracellular metabolic flux is isotope tracing coupled with computational metabolic flux analysis. This has become a central technique in studies of cancer cellular metabolism. A fundamental limitation in our understanding of cellular metabolism is having no information on cell-cell variability in metabolic activity. Standard isotope tracing techniques are applied on a population of cells that are heterogeneous in terms of cell-cycle phase, providing no information on alterations in metabolic flux throughout the cell-cycle - practically aiming to estimate the ¡°average¡± flux through the cell population. Here, we developed a temporal-fluxomics approach to derive a comprehensive and quantitative view of alterations in metabolic fluxes throughout the mammalian cell cycle. This is achieved by combining pulse-chase LC-MS based isotope tracing in synchronized cell populations with computational deconvolution and metabolic flux modelling. We find that TCA cycle fluxes are rewired as cells progress through the cell cycle with complementary oscillations of glucose versus glutamine-derived fluxes: Oxidation of glucose-derived flux peaks in late G1 phase while oxidative and reductive glutamine metabolism dominates S phase. These complementary flux oscillations maintain a constant production rate of reducing equivalents and oxidative phosphorylation flux throughout the cell cycle. The shift from glucose to glutamine oxidation in S phase plays an important role in cell cycle progression and cell proliferation.

  • CGGC Seminar: From Geometry to Simulation and Back: Numerical Design in Primary Manufacturing Processes

    דובר:
    סטפני אלגטי (אונ' אאכן)
    תאריך:
    יום רביעי, 5.9.2018, 13:30
    מקום:
    חדר 337, בניין טאוב למדעי המחשב

    Using a mold or die, primary shaping manufacturing processes form material from an initially unshaped state (usually melt) into a desired shape. Examples of such a process are extrusion or high-pressure die casting — processes that are responsible for many products of our everyday life, from pipes to yoghurt cups. From the design perspective, what these processes have in common is that the exact design of the mold cannot be determined directly and intuitively from the product shape. This is due to the non-linear behavior of the material regarding the flow and solidification processes. Consequently, shape optimization as a means of numerical design can be a useful tool in mold development.

    The core of our optimization tool is the in-house flow solver XNS, which combines a space-time method with either polynomial or isogeometric shape functions with a GLS stabilization. XNS is able to exploit the common communication interfaces for distributed-memory systems. The flow solver has been coupled with the open-source optimization frameworks NLOPT and Dakota. For geometry representation, we utilize an in-house spline library which supports both NURBS and T-splines. Spline representations are very natural in engineering design, as they allow the shape optimization result to be easily transferred back into the CAD-based design process. Furthermore, they require a low number of optimization parameters and allow the incorporation of manufacturing constraints. Isogeometric analysis aligns well with this type of shape-optimization.

    Topics discussed will be our approach to shape optimization as well as methods for simulating the flow through, in and behind the mold/die. The importance of the geometrical respresentation and the resulting challenges in this area will be emphasized.

  • בית-ספר קיץ השביעי בנושא אבטחת סייבר

    The 7th Summer School on Cyber and Computer Security

    תאריך:
    יום שלישי, 2.10.2018, 09:30
    מקום:
    טכניון

    מרכז המחקר לאבטחת סייבר ע"ש הירושי פוג'יווארה יקיים את בית-ספר קיץ השביעי על אבטחת סייבר ומחשבים: "Trusted Execution and Hardware Side Channels"

    הכנס יתקיים בימים ג'-ה', 4-2 באוקטובר, 2018, בטכניון, חיפה.

    מארגני הכנס:
    פרופ' מרק זילברשטיין  – טכניון
    פרופ' יוסי אורן
     – אוניברסיטות בן-גוריון

    משתתפים:
    Christof Fetzer, TU Dresden
    Daniel Genkin, University of Michigan
    Herbert Bos, VU Amsterdam
    Ittai Anati, Intel Israel
    Taesoo Kim, Georgia Tech

    ההרשמה תיפתח ב-2.9.2018.

    פרטים נוספים ומידע על מרכז המחקר לאבטחת סייבר ע"ש הירושי פוג'יווארה.

  • Predicting a Better Future for Asynchronous Stochastic Gradient Decent with DANA

    דובר:
    עידו חכימי, הרצאה סמינריונית לדוקטורט
    תאריך:
    יום שלישי, 30.10.2018, 14:30
    מקום:
    טאוב 601
    מנחה:
    Prof. Assaf Schuster

    Distributed training can significantly reduce the training time of neural networks. Despite its potential, however, distributed training has not been widely adopted due to the difficulty of scaling the training process. Existing methods suffer from slow convergence and low final accuracy when scaling to large clusters, and often require substantial re-tuning of hyper-parameters.

    We propose DANA, a novel approach that scales to large clusters while maintaining state-of-the-art accuracy and converge speed without having to re-tune parameters that are optimized for training on a single worker. By adapting Nesterov Accelerated Gradient to a distributed setting, DANA is able to predict the future position of the model's parameters and so mitigate the effect of gradient staleness, one of the main difficulties in asynchronous SGD.