Ryan Mcbride Email & Phone Number

Surrey, Bc, Ca

I have taught CMPT 339 (Introduction to Databases) in Fall 2020 and am in the process of teaching an Artificial Intelligence course at Trinity Western University this Spring. The focus of my teaching is to present the technical beauty of Computing science and societal implications of these technologies, whether in how data storage changed the course of history (e.g., management of COVID compared to the Spanish Flu) to the dangers of Artificial Intelligence in reinforcing problematic beliefs via filter bubbles.

Langley, British Columbia, Canada

At Trinity Western University, I am working with Professor Herbert Tsang on software to predict what RNA sequences will fold into certain structures of biological interest. My role involves exploring potential extensions using AI techniques, such as surveying more diverse Simulated Annealing schedules for the problem. Some of this ongoing research was accepted to an Evolutionary Algorithm conference, IEEE CEC 2020.

Surrey, British Columbia, Canada

Link: https://www.rom2consulting.com/The goal of ROM2 consulting is to bring data science, software and apps to augment decision-making from data for use by non-profits, especially in British Columbia’s Lower Mainland. Our team’s background combines major data science projects, including models to prevent powers outages in British Columbia to guide the most useful distribution of repair crews and models to predict and remove carcinogenic electrical devices, experience in reviewing data science publications at international data science conferences, and in volunteering work to consult and guide non-profits with fundraisers and grant applications.

Burnaby, British Columbia, Canada

As a sessional instructor, I taught Discrete Math I, an undergraduate course that covers the foundations of logic, set theory, and other mathematical principles for Computing Science Majors at Simon Fraser University during a pandemic. I presented and designed lecture notes with an emphasis on how a strong mathematical and logical foundation helps students better understand the world, society, and modern data science. For example, I discussed ideas of universal quantified statements (e.g., every Quaker is a pacifist) in terms of how they represent universal laws such as gravity or ``this security protocol is correct’’ where a single counterexample can disprove it resulting in a paradigm shift of invalidating a whole theory. For existential quantifiers, I discussed how hard it is to disprove conspiracy theory based claims, such as that ``somebody must live on the moon’’ and societal issues of easily getting trapped in filter bubbles of ignoring material that goes against existing views and prejudices.I also designed assignments, midterms, and tutorial exercises to judge a student’s breadth and depth of understanding of the concepts and managing three teaching assistants in fair and consistent marking. My questions emphasize real world applications, such as having questions related to explaining discrete math concepts to non-experts in business contexts. And a couple funny student reviews I got: “Thank you for making Discrete Math somehow enjoyable” and “YAAAIII MCBRIDE YOU’RE MY HOMEBOY.”

Data-Mining Lab

In collaboration with BC Hydro and my senior supervisor, I developed new data analysis prediction models to rank which electrical networks were most at-risk of failing during forecasted storms. My contributions on this project include: Innovative Data Science Solutions: From discussions with BC Hydro, I developed the idea of Cost-Sensitive Learning to Rank, a new Machine Learning problem of building a model to allocate resources to maximally prevent as much outage damage or other notions of damage as possible. This model was determined novel in an extensive comparison to existing research in ranking problems, built for problems such as returning relevant pages on Google. Experimental Validation: I developed a suite of methods to adapt the majority of existing ranking methods, using Deep Learning and other techniques, to this new problem; in experiments on historical outage data, forest fire prevention, and quality assurance of batches of concrete, these methods prevented over 20% more damage than state-of-art competitors. Publication to AAAI 2019: This work was accepted in the research track of AAAI 2019, a top-10 machine learning conference, as a paper, an oral presentation, and a poster; less than 16.2% of submissions were accepted. Acceptance to such a conference generally implies that the research is useful to the general population of data scientists due to peer review standards.

I was a teaching assistant for CMPT 454 Database Systems II (an undergraduate course regarding technology used to make SQL and other database queries efficient) in the summer 2018 where I marked hundreds of assignments/exams, contributed to the course direction by providing assignment/exam questions, and answered student queries and concerns.Some choice reviews from students: "Willingness to help. No weakness. He is great already, no need to improve :)", "Ryan frequently answers Piazza questions [note: the online forum to discuss course issues] and does so diligently. He provided lots of feedback for assignments", and "Keep up the good work bud!".

I developed Machine Learning models to help BC Hydro identify electrical transformers that should be tested for certain carcinogenic toxins, resulting in: Innovative Data Science Solutions: Collaborating with my supervisor, BC Hydro, and an undergraduate student that I mentored, we developed a new machine learning problem that was more consistent with how companies deal with these risk management problems in practice. I developed more suitable models for this problem and even a better statistical probability estimation technique. Experiment Validation: These new models, implemented in Java, outperformed competitors: for example, our method only unnecessarily tested 10% of safe transformers when the best competitor would extraneously test 90% of such objects, incurring excessive cost. Experiments on other data sets, such as identifying non-cancerous patients, showed similar improvements. Publications: This work was a regular paper at the 2014 IEEE International Conference on Data-Mining (ICDM), a top data-mining conferences where less than 10% of the 727 submissions were accepted, as well as at SIAM Data-Mining 2017, another top-tier data-mining conference (<26% of papers were accepted).

Simon Fraser University

The three NSERC awards were with the computational logic lab to develop:- Software to represent and search for flaws in cryptographic protocols (culminates in that "Protocol Verification in Theory of Action" paper), work was with Jim Delgrande.- Research to explore the relationship between logical satisfiability formula's and their tree-width (work still ongoing with Dave Mitchell).

Simon Fraser University

Simon Fraser University

Doctor Of Philosophy (Science), Computer Science, 4.330/4.330

Bachelor Of Applied Science (B.A.Sc.), Computer Science, 3.930/4.330

Honours in Computing Science with Distinction (>3.50 GPA) along with a Minor in Mathematics.