| slug | title |
|---|---|
past |
Past Events |
import sep252023 from './sep252023.jpg'
Our CREATE SE4AI community came together on September 25th, 2023 for the first Trainee Talks of the fall semester. We heard Queen's Master's candidate Divya Kamath present her thesis research on Pragmatic Approaches to Schedule Less Builds in Continuous Integration. Then, Mohayeminul Islam, PhD student from University of Alberta shared his work on Characterizing Python Library Migrations.
The video is now available to view on our CREATE SE4AI YouTube channel.
import apr252023 from './apr252023.jpg'
CREATE SE4AI is pleased to present Patrick Mesana, Consultant at National Bank of Canada and PhD Candidate at HEC, who spoke on Balancing Data Valorization and Privacy: Industry Challenges and Opportunities.
In his talk, Patrick delved into the distinct characteristics of data as a digital asset and examined the challenges of striking a balance between data valorization and privacy. Emphasizing the importance of data management in large organizations, such as financial institutions, he addressed the emerging challenges posed by privacy laws and the rapid growth of AI. Through a discussion of evolving practices like privacy-by-design, advances in privacy engineering, and innovative privacy technologies such as data synthesis, he offered valuable insights for navigating the intricate future of data privacy within organizations. Furthermore, he brought to light potential future disruptors, such as open banking in Canada, particularly relevant to the finance sector.
About the Speaker:
Patrick Mesana began his career as a software engineer, gaining experience with consulting companies and startups before transitioning into the field of Data Science. As a former manager in the Data Office of National Bank of Canada (NBC), Patrick encountered numerous data management challenges. He is currently pursuing a PhD in decision science to explore the tradeoffs between data privacy and data value. In addition, Patrick serves as a researcher and scientific lead on these topics for the AI Factory at NBC.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
import april32023 from './april32023.jpg'
The Spring Trainee Talks webinar was a huge success and we were pleased to have 27 participants join!
In this webinar, 2 PhD students from Polytechnique Montréal under Prof. Foutse Khomh delivered their research topics. Vahid Majdinasab spoke on his topic of Mutation Testing of Deep Reinforcement Learning Based on Real Faults, followed by Forough Majidi, who shared her Empirical Study on the Usage of Automated Machine Learning Tools. A 10 min Q&A session followed after each presentation.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
On March 18th , CREATE SE4AI took part in WIE Inspire WIE Empower, an annual event hosted by the Women in Engineering (WIE) student association at the Gina Cody School of Engineering and Computer Science at Concordia University. Through their various outreach and mentorship activities, WIE encourages young women to explore engineering and computer science as career options.
This year, WIE welcomed over 80 high school girls aged 12-17 to Concordia for an exciting day of inspiration and exploration. After a welcome address by Dr. Gina Cody, students participated in 4 hands-on, industry-led workshops. CREATE SE4AI trainees and Master’s students Divya Kamath (Queen’s University), Lorena Barreto (Concordia University) and Khaled Badran (Concordia University), along with Program Coordinator Lori Akiyama developed and delivered a fun, interactive workshop on AI and Chatbots to 4 groups of students. The goals of the workshop were to:
- Introduce the basic concepts of AI and chatbots to the students
- Encourage the students to develop confidence interacting with technology
- Stimulate a discussion of ethics of AI in society through a group reflection exercise
At the end of the workshop, the students shared their experiences interacting with chatbots and reflected on the importance of creating responsible and ethical AI.
CREATE was proud to contribute to this important initiative, and we look forward to more opportunities to engage with youth and inspire them to learn more about software engineering and AI in the future.
import wie1 from './wie1.jpg' import wie2 from './wie2.jpg' import wie3 from './wie3.jpg' import wielogo from './wielogo.jpg'
import feb272023 from './feb272023.jpg'
In the first Trainee Talks webinar of 2023, Pierre-Olivier Côté, a Master's student at Polytechnique Montréal spoke about Quality Issues in Machine Learning Software Systems. Immediately following, University of Alberta PhD student Akalanka Galappaththi presented his work on “A Data Set of Generalizable Python Code Change Patterns”.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
import feb102023 from './feb102023.jpg'
Deploying Effective Machine Learning Models in Production: Two Use Cases
Dr. Gabor Melli, VP of Engineering, AI/ML at San Francisco-based Medable joined us for our first industry talk webinar of 2023! Gabor Melli is a seasoned professional with a proven track record of leading teams and driving positive change within organizations. His background includes implementing evidence-based solutions and facilitating change at large enterprises such as Sony PlayStation, Microsoft, AT&T, T-Mobile, and Wal*Mart, as well as at start-ups including Datasage, Meals.com, VigLink, OpenGov.com, and Medable.com. Gabor has also presented his research at various applied AI research conferences, including AAAI, KDD, LREC, and ICDM.
Abstract
Predictive machine learning has the ability to automate and optimize workflows across a wide range of industries. In this session, we will explore real-life examples of how predictive machine learning is being used at Sony Interactive Entertainment (SIE) and Medable. At SIE, we will delve into their scalable, real-time, low-latency predictive ML-based solutions that are implemented on the cloud in order to personalize purchasing and playing experiences. Medable, on the other hand, has implemented an automated, decentralized clinical trials platform on the cloud in order to greatly improve efficiency and accuracy. Throughout this session, we will also discuss the role of machine learning engineering and the exciting opportunities that lie ahead in this field. Overall, this session aims to provide a comprehensive overview of the various applications and potential of predictive machine learning.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
import trainee_talk_2 from './trainee_talk_2.jpg'
Trainee Talks is a monthly webinar series featuring CREATE SE4AI trainees sharing their research topics in software engineering and AI.
This time, Queen's University Master's student Ernesto Lang Oreamuno joined us and spoke about Documentation Practices of Third-Party Models and Datasets.
Immediately following, Rached Bouchoucha, École Polytechnique Master's student presented his work on de-bugging Deep Reinforcement Learning (DLR) Frameworks.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
import telecom2 from './telecom2.jpg'
CREATE SE4AI is pleased to present AI Adoption in Telecom Networks as part of our Industry Talks webinar series.
Did you know that revenue generated from AI automation in telecom networks is expected to reach USD 200 billion by 2027? What are some of the current challenges Ericsson, a global leader in ICT solutions faces as they adopt new technologies in AI and ML in their products and services, and how are they overcoming them?
In his presentation, Karthikeyan Premkumar, Data Scientist at Ericsson used case studies to highlight several touchpoints involving the ML model development process, tools, data quality, ML Ops (monitoring and retraining) and deployment as part of model industrialization.
About the Speaker
Karthik Premkumar has 18 years of experience in the telecom industry. His expertise lies in industrializing AI/ML solutions, designing system and solution architectures in BSS, Cloud and infrastructure platforms for telecom networks.
In his current role as Data Scientist at Ericsson in Montreal, QC, Karthik Premkumar designs knowledge models and machine reasoning technologies for cognitive networks. He has published over 10 patents and papers on telecom analytics.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
Topics: Dependency Management and Testing of AI-based Software Systems
Our monthly CREATE SE4AI Trainee Talks webinar series kicked off with 2 graduate Software Engineering students presenting their research.
In the first of our series of trainee-led webinars, we invited 2 graduate students to deliver their research topics to the broader public.
Ahmed Haj Yahmed, a Master's student at Polytechnique Montréal presented DiverGet: a Search-based Software Testing Approach for Deep Neural Network Quantization Assessment. Immediately following, Concordia University PhD candidate Jasmine Latendresse presented her work on software production dependencies which she recently delivered at the IEEE/ACM 2022 Conference on Automated Software Engineering (ASE '22) in Michigan.
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
Our 1st annual CREATE SE4AI Program Retreat is a wrap! It was an inspiring day of celebrating our collective accomplishments and exchanging ideas amongst an amazing team of students, professors and industry partners, live and in person at the Chateau Vaudreuil. The future is so bright for this group of talented trainees, they’re gonna need shades…
import retreat from './retreat-group.jpg' import brainstorm from './brainstorm.jpg' import group1 from './group1.jpg' import group2 from './group2.jpg' import group3 from './group3.jpg' import group4 from './group4.jpg'
CREATE SE4AI trainees, profs and partners enjoyed a webinar presented by Olivier Blais, co-founder and VP of Decision Science at Moov AI in Montréal. Olivier spoke about "Delivering High Quality Machine Learning Models" including:
import olivier from './olivier.jpg'
-
What it means to develop high quality artificial intelligence
-
A better approach to ML model evaluation
-
Existing and future certifications for AI systems
Bio: https://moov.ai/en/olivier-blais/
The webinar recording is available to view on our CREATE SE4AI YouTube channel.
CREATE SE4AI hosted a webinar by Sumon Biswas, PhD Candidate at Iowa State University on "Understanding and Reasoning Fairness of Machine Learning Pipeline".
Link to presentation: https://www.youtube.com/channel/UCjCS6a_K301Ocg9z5Qd1GWA/videos
About the Speaker
import sumon from './sumon-biswas.jpg'
Sumon Biswas is a Computer Science Ph.D. candidate at Iowa State University (ISU) and a Research Assistant in Laboratory for Software Design at ISU under the supervision of Professor Hridesh Rajan. His research interests are in the intersection of Software Engineering, Programming Languages, and Artificial Intelligence. He has worked on Machine Learning (ML) software repository mining and analysis in large-scale using the Boa framework. He worked on building Python language support for Boa to analyze ML programs and Jupyter Notebooks. Currently, he is working in the D4 (Dependable Data-Driven Discovery) initiative at ISU towards increasing the dependability of data-driven software. Specifically, he is conducting research on understanding the societal bias in ML models and reasoning about fairness property and its mitigation in ML pipelines. His research results appeared in reputed software engineering venues including ICSE and ESEC/FSE.
Homepage: https://sumonbis.github.io