The following is the schedule of upcoming department seminars. Seminars will be added to the schedule throughout the year. All seminars are free of charge and are open to all members, affiliates, and colleagues of the university community. Seminars are typically held on Tuesday or Thursday at 1:00 pm. If you would like additional information or if you might be interested in presenting a seminar, please contact Helen Cameron or the Department of Computer Science.
Departmental seminar - Ji Hyun KoWhen: March 02, 2017 @ 1:00pm
Where: EITC E2-165
Title: How can we improve early diagnosis of Alzheimer’s disease using machine learning and brain imaging?
Alzheimer’s disease (AD) is the most prevalent cause of dementia that affects >747,000 Canadians as of 2011. Number of disease modifying therapies and preventive interventions are under development. Early detection of AD may benefit patients and caregivers by providing a better chance of benefiting from treatment, relief from anxiety about unknown problems, more time to plan for the future and early access to psychosocial education for both patients and caregivers.
The current consensus is that patients with mild cognitive impairment (MCI), a state referring with cognitive deficits but not serious enough to interfere with patient’s independence, are at higher risk of developing AD or other dementia than age-matched normal controls. Therefore, routine tests of mental/neuropsychological performance and interviews by physicians may be used for early detection of AD. However, only 20-40% of amnestic MCI patients eventually convert to AD. The use of [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET), a brain imaging technique that estimate regional glucose metabolism, has been widely accepted to aid physicians to complement clinical diagnosis, but the subjective impressions of FDG-PET readings are often found to be equivocal. Therefore more objective quantification of risk estimate using FDG-PET is highly demanded.
Here, we are investigating the use of scaled subprofile modeling (SSM), a form of principal component analysis (PCA), to automate the FDG-PET readings. Early preliminary data shows promising results. We are currently trying to employ machine learning algorithm to further improve the performance of automated early diagnosis of AD.
Departmental seminar - John Craig Muller, Rajendra Singh and Eric MellickWhen: March 07, 2017 @ 1:00pm
Where: EITC E2-105
Title: Use of Parallel and High Performance Computing in Power Systems Simulation and Its Challenges
Abstract:Parallel processing is becoming a larger and more important area of study in both academic research and within practical application. Most Power system simulation software utilize parallel processing. Simulations in power systems are a collection of smaller electrical networks, complex models and switching elements, which run iteratively at very small computation steps. Networks are solved and may require to exchange values at the end of each step. With the increase in the size of network and details in models, computation becomes complex and intensive. Thus, use of advanced computing becomes evident. Data parallel processing increases the throughput of simulation studies by utilizing significantly large number of computing units to solve multiple data sets in parallel using a single executable. Initial parameters are fed to the simulation and the final result is collected, thus incurring low Inter-Process communication (IPC). Task parallel processing in simulation software is used to improve the performance of a single simulation.The challenge is to realize a very large scale simulation into multiple simulations/tasks and then connected with each other using an IPC mechanism to exchange values. Values are exchanged every computation step, for example a 10 second simulation running at 10 us computation step requires 1 million message exchanges. This makes task parallelism a very communication intensive approach. In addition the scale of the system makes it exceedingly expensive to build on a real-time hardware based solution and still retain full fidelity desired.IPC mechanisms such as local memory have low communication latencies and perform well in localhost. TCP/IP is used for local and distributed computing but incurs high communication latencies.InfiniBand network infrastructure and Remote Direct Memory Access (RDMA) over Converged Ethernet (RoCE) provides latencies across hosts in the range of 1-2 us. Simulation software can be made RDMA/RoCE aware to achieve unmatched performances by utilizing scalability of distributed computing yet achieving local memory based communication latencies.A problem generally faced in task parallel approach is breaking apart a large, tightly coupled power system simulation into independent tasks. Due to the large number of inter-dependencies in a mesh configuration, it is very difficult to determine the optimal break points. Each break reduces the computational burden in the individual subnetworks, however these breaks add to the number of processors and interconnections required. Since large systems can be comprised of hundreds of sub-networks and potentially hundreds of inter-connects, the process of optimally mapping them onto a fixed number of processors is challenging. Identifying the break points may require an intelligent method using graph theory to quantify the influence of each subnetwork.1) Each process is optimal when evenly loaded, while minimizing inter-process communications.2) Each machine must be maximally loaded to take advantage of local processing capabilities.Both of these load balancing challenges can be represented by a graph partitioning problem, where given a graph and a number of required partitions, an optimal set of partitions can be obtained, knowing in advance the target hardware platform.
Departmental Seminar: Md. Altaf-Ul-AminWhen: March 21, 2017 @ 1:00pm
Where: EITC E2-105
Title: Structural Similarity based Multifaceted Analysis of Metabolites using KNApSAcK Database and DPClus Algorithm
Abstract:Initially, we developed KNApSAcK as a species-metabolite relational database and subsequently, inspired by its popularity we extended it to KNApSAcK family databases by adding different types of omics data together with data regarding edible plants and traditional medicines mainly focusing human health care and ecology. Previously we also developed graph clustering algorithms DPClus and DPClusO, which we and many other researchers applied to analysis of versatile omics data. In the present talk, first, I will briefly focus on the KNApSAcK database and the DPClus algorithm. Then I will discuss two of our recent publications. One is on development and mining of a volatile organic compound (VOC) database. VOCs play important roles in chemical ecology specifically in the biological interactions between organisms and ecosystems. VOCs are also important in the health care field as they are presently used as biomarkers to detect various human diseases. The other topic of my talk is a network-based approach to analyze relationships between 3D structure and biological activity of metabolites. By applying our method to a data set consisting of 2072 secondary metabolites and 140 biological activities reported in KNApSAcK Metabolite Activity DB, we have obtained 983 statistically significant structure group-biological activity pairs which implies relations between chemical structures and biological activities of metabolites.Bio:Md. Altaf-Ul- Amin received B.Sc. degree in Electrical and Electronic Engineering from Bangladesh University of Engineering and Technology (BUET), Dhaka, M.Sc. degree in Electrical, Electronic and Systems Engineering from Universiti Kebangsaan Malaysia (UKM) and PhD degree from Nara Institute of Science and Technology (NAIST), Japan. He received the best student paper award in the IEEE 10 th Asian Test Symposium. Also, he received two other best paper awards as a co-author of journal articles. He previously worked in several universities in Bangladesh, Malaysia and Japan. Currently he is working as an associate professor in Computational Systems Biology Lab of NAIST. He is conducting research on Network Biology, Systems Biology, Cheminformatics and Biological Databases. He published around 60 peer reviewed papers in international journals and conference proceedings.
Departmental Seminar - Catherine Westlake and Kent BrewerWhen: April 04, 2017 @ 1:00pm