Events

Jun
20
Thu
Fields-CQAM Public Lectures: What is missing from common practice in machine learning? @ Carleton University
Jun 20 @ 19:00 – 20:00

Fields-CQAM Public Lectures: Ali Ghodsi, University of Waterloo

 

What is missing from common practice in machine learning?

AI, and machine learning in particular, is enjoying its golden age. Machine learning has changed the face of the world over the past two decades but we are still a long way from achieving a general artificial intelligence. In this talk, I will discuss a couple of elements that I believe are missing from common practice in machine learning, including incorporating causality and creating a new framework for unsupervised learning.

 

Biography

 

Ali Ghodsi is a Professor in the Department of Statistics and Actuarial Science at the University of Waterloo. His research involves statistical machine-learning methods. Ghodsi’s research spans a variety of areas in computational statistics. He studies theoretical frameworks and develops new machine learning algorithms for analyzing large-scale data sets, with applications to bioinformatics, data mining, pattern recognition, robotics, computer vision, and sequential decision making.

DATE:

THURSDAY, JUNE 20TH, 2019.

PRESENTATION

6:00 PM – 7:00 PM.

NETWORKING

7:00 PM – 8:00 PM.

LOCATION

HEALTH SCIENCE BUILDING, RM. 1301 (LOCATED ON THE GROUND FLOOR), CARLETON UNIVERSITY.

FREE ADMISSION FOR THIS PUBLIC LECTURE.
PLEASE REGISTER HERE.

Jun
21
Fri
FIELDS CENTRE OF QUANTITATIVE MODELLING AND ANALYSIS: WORKSHOP ON Machine Learning in the Presence of Class Imbalance @ Residence Commons, Carleton University
Jun 21 @ 08:30 – 16:30
FIELDS CENTRE OF QUANTITATIVE MODELLING AND ANALYSIS: WORKSHOP ON Machine Learning in the Presence of Class Imbalance @ Residence Commons, Carleton University | Ottawa | Ontario | Canada

 

8:30 am – 9:00 am Registration
9:00 am – 9:15 am Opening Remarks Rafik Goubran Carleton University
9:15 am – 10:00 am Keynote Presentation:

Data Mining and Machine Learning for Authorship and Malware Analyses
Abstract

Benjamin C. M. Fung
Biography
McGill University
10:00 am – 10:30 am Break
10:30 am – 11:45 am Cybersecurity: Top 5 class imbalance ML challenges and data sets
Abstract
Stephan Jou
Biography
Interset
Class Imbalance in Fraud Detection
Abstract
Robin Grosset
Biography
MindBridge Analytics Inc.
Handling class imbalance in natural language processing
Abstract
Isuru Gunasekara
Biography
IMRSV Data Labs
11:45 am – 12:45 pm Lunch
12:30 pm – 2:10 pm Adaptive learning with class imbalanced streams
Abstract
Herna L. Viktor
Biography
University of Ottawa
Radar-based fall monitoring using deep learning
Abstract
Hamidreza Sadreazami
Biography
McGill University
Privacy-preserving data augmentation in medical text analysis
Abstract
Isar Nejadgholi
Biography
National Research Council
Failure modelling of a propulsion subsystem: unsupervised and semi-supervised approaches to anomaly detection
Abstract
Julio J. Valdés
Biography
National Research Council
2:10 pm – 2:25 pm Break
2:25 pm – 3:40 pm TBD Reddy Nellipudi DB Schenker
AuditMap.ai: Hierarchical Sentence Classification in Unstructured Audit Reports
Abstract
Daniel Shapiro
Biography
Lemay.ai
Deep Learning techniques for unsupervised anomaly detection
Abstract
Dušan Sovilj
Biography
RANK Software Inc.
3:40 pm – 3:50 pm Closing Remarks

 

Oct
11
Fri
Advanced optical sources for spectrally efficient photonic systems – Liam Barry, Dublin City University @ Advanced Research Complex (ARC), uOttawa
Oct 11 @ 09:00 – 10:30

Advanced Optical Sources for Spectrally Efficient Photonic Systems
Liam Barry,
Dublin City University

 

Abstract

The continuing growth in demand for bandwidth (from residential and business users), necessitates significant research into new advanced technologies that will be employed in future broadband communication systems. Two specific technologies which are becoming increasingly important for future photonic
systems are wavelength tunable lasers and optical frequency combs. Although these topics have been studied for over two decades their significance for the development of future ultra-high capacity photonic systems has only recently been fully understood. Wavelength tunable lasers are currently becoming the
norm in optical communication systems because of their flexibility and ability to work on any wavelength. However, as their operating principles are different to standard single mode lasers they can effect how future systems will operate.

For example as optical transmission systems move towards more coherent transmission (where the data is carried using both the intensity and phase of the optical carrier), the phase noise in these tunable lasers will become increasingly important. Optical frequency combs also have many applications for
future photonics systems, and for telecommunications they can be used to obtain the highest spectral efficiency in optical transmission systems by employing the technology of optical frequency division multiplexing (OFDM) that has been widely employed to increase spectral efficiency in wireless systems. Wavelength tunable lasers and optical frequency combs are thus topics at the leading edge of current photonics systems research, and their detailed understanding promises new applications in all-optical signal processing, optical sensing and metrology, and specifically telecommunications. This talk will focus on the development and characterization of various wavelength tunable lasers and optical frequency combs, and then outline how these sources can be employed for developing optical transmission systems and networks which make the best use of available optical spectrum.

Bio

Liam Barry received his BE (Electronic Engineering) and MEngSc (Optical Communications) from University College Dublin and his PhD from the University of Rennes. His main research interests are: all-optical signal processing, optical pulse generation and characterization, hybrid radio/fibre communication
systems, wavelength tuneable lasers for reconfigurable optical networks, and optical performance monitoring. He has worked as a Research Engineer in the Optical Systems Department of France Telecom’s Research Laboratories (now known as Orange Labs), and a Research Fellow at the Applied Optics Centre in Auckland University. He is currently a Full Professor in the School of Electronic Engineering at Dublin City University, establishing the Radio and Optical Communications Laboratory, and is a Principal Investigator for Science Foundation Ireland. He has published over 500 articles in internationally peer reviewed journals and conferences, holds 9 patents in the area of optoelectronics, and has co-founded two companies in the photonics sector.

 

Oct
19
Sat
IEEE Ottawa Seminar Series on AI and Machine Learning – Sponsored by IEEE Ottawa CS Chapter, ComSoc Chapter, and SP Chapter, jointly with Vitesse- Reskilling
Oct 19 @ 00:07 – 01:07

Date Wednesday, Oct 30, 2019

Location 359 Terry Fox Drive, Kanata, Ontario

Agenda

       11:30 AM – 12:00 PM: Light Lunch and Networking

       12:00 PM – 1:00 PM  : Presentation and Q&A

1:00 PM – 1:30 PM    : Post Presentation Networking

Title of the Talk AI-Powered 5G Networks
& Beyond

Speaker  Hatem Abou-zeid 

Summary

5G Networks are anticipated
to transform modern societies by providing an ultra-reliable, high-speed
communications infrastructure that will connect billions of devices including
vehicles, machines, and sensors. Both the complexity of such networks and the
diversity of application requirements will be unprecedented. This mandates
novel, autonomous network configuration and operation that can anticipate and
react to changes in traffic, topology, and interference conditions to ensure
seamless quality of experience and reliability. In this talk I will discuss
AI-driven networking use-cases elaborating on the practical challenges of
industrial deployments. I will then highlight directions where research is
needed to further expedite and facilitate the development of AI-powered
networks.

Biography

Hatem Abou-zeid is a
Senior 5G Systems Designer at Ericsson Canada where he drives research and
system development for 5G radio access networks. Prior to that he held
industrial positions at CISCO Systems and Bell Labs in addition to postdoctoral
and research assistant affiliations at Queen’s University, Canada. His research
focuses on the application of machine learning in 5G networks with particular
emphasis on anticipatory and adaptive algorithms drawing on methods from
reinforcement learning, spatio-temporal forecasting, deep learning and
stochastic optimization. Dr. Abou-zeid is very passionate about developing
strong industry-university collaborations that foster applied, innovative
research, and he leads multiple academic partnerships on intelligence and
analytics in future networks.

Dec
3
Tue
Advanced semiconductor lasers: Ultra-low operating energy and heterogeneous integration with Si photonics devices @ University of Ottawa, Room 223
Dec 3 @ 13:00 – 14:00

IEEE Photonics Society Distinguished Lecturer Program

Advanced semiconductor lasers:Ultra-low operating energy and heterogeneous integration with Si photonics devices

Shinji Matsuo, NTT Photonics Laboratories, Japan

Abstract: The electrical power consumed in data transmission systems is now hampering efforts to further increase the speed and capacity at various scales, ranging from data centers to microprocessors. Optical interconnects employing an ultralow energy directly modulated lasers will play a key role in reducing the power consumption. Since a laser’s operating energy is proportional to the size of its active volume, developing high-performance lasers with a small cavity is important. For this purpose, we have developed membrane DFB and photonic crystal (PhC) lasers, in which active regions are buried with InP layer. Thanks to the reduction of cavity size and the increase in optical confinement factor, we have achieved extremely small operating energy and demonstrated 4.4-fJ/bit operating energy by employing wavelength-scale PhC cavity. Reduction of the cost is also important issue because huge number of transmitters are required for short distance optical links. For this purpose, Si photonics technology is expected to be a potential solution because it can provide large-scale phonic integrated circuits (PICs), which can reduce the assembly cost compared with transmitters constructed by discrete devices. Therefore, heterogeneous integration of III-V compound semiconductors and Si has attracted much attention. For fabricating these devices, we have developed wafer-scale fabrication procedure that employs regrowth of III-V compound semiconductors on directly bonded thin InP template on SiO2/Si substrate. A key to realize high-quality epitaxial layer is total thickness, which must be below the critical thickness, typically 430 nm. Thus, membrane structure is quite suitable for heterogeneous integration. I will talk about our recent progress, focusing on ultralow-powerconsumption directly modulated lasers and their photonic integrated circuit. I will also describe progress in heterogeneous integration of these lasers and Si photonics devices.

Bio: Dr. Matsuo received a B.E. and M.E. degrees in electrical engineering from Hiroshima University, Hiroshima, Japan, in 1986 and 1988, and the Ph.D. degree in electronics and applied physics from Tokyo Institute of Technology, Tokyo, Japan, in 2008. In 1988, he joined NTT Optoelectronics Laboratories, Atsugi, where he was engaged in research on photonic functional devices using MQW-pin modulators and VCSELs. In 1997, he researched optical networks using WDM technologies at NTT Network Innovation Laboratories, Yokosuka. Since 2000, he has been researching InP-based photonic integrated circuits including fast tunable lasers and photonic crystal lasers at NTT Photonics Laboratories, Atsugi. Dr. Matsuo is a member of the IEEE Photonics Society, Japan Society of Applied Physics and the Institute of Electronics, Information and Communication Engineers (IEICE) of Japan.

IEEE Ottawa Section Logo

© Copyright 2020 IEEE – All rights reserved. Use of this website signifies your agreement to the IEEE Terms and Conditions.

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity.