Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

McMaster University

$41,000.00

Towards a Robust and Trustworthy Machine Learning for Visual Contents in Data-centric Applications

EF 2023-24

Résumé
Recently, machine learning (ML) performance has drastically increased as a result of powerful computing facilities and large datasets. However, the inherent weaknesses in the reliability and security of data-centric machine learning applications may result in performance degradation, system misbehavior, and/or privacy breach. Building trustworthy data-centric learning systems is among the biggest and most impactful challenges in the next decade. My goal is to enhance the understanding of the vulnerabilities and/or biases caused by imperfect training sets and to develop the fundamental tools, metrics and methods to explain and mitigate them. We seek to formulate the trustworthy machine learning problems and design solutions to ensure the robustness, trustworthiness, and privacy of data-centric machine learning systems. This research program focuses on developing practical approaches to build dependable and trusted models/systems with noisy training data (which contains non-negligible portion of mislabeled samples), manipulated data, biased data, and inappropriate scale of data, etc. One step further, the research outcomes will be used to algorithmically and automatically generate reliable training sets from social media apps, webs search and outsourcing, and/or identify the mislabeled training samples from the third part datasets. In addition, the proposed research will address privacy-preserving image data sharing and publishing. This will allow machine learning applications to efficiently get the "good" data samples in an appropriate scale, therefore yielding reliable models. It is anticipated that the proposed research will result in profound reliability, security, and privacy implications when we deploy ML systems and help to shift the ML paradigms from model-centric to data-centric. The research agenda will attract participation from a diverse and inclusive team of undergraduate and graduate students, postdoctoral fellows, and industry personnel. They will be sought for their unique expertise in AI, multimedia big data analysis, security & privacy, and social networks. The research outcomes will directly benefit the Canadian society and economy by providing small companies (who don't have enough expertise, computing facilities, and data) an opportunity to use AI technologies successfully, by enhancing data quality and model reliability for large companies, and by leveraging the Canadian leadership in artificial intelligence in terms of dependable, high-performance and low-cost features.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
RGPIN-2023-04766
Name nom
He, Wenbo
Department département
Computing and Software
Organization id
27
Institution établissement
McMaster University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
41000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
9999
Research subject group en
Not available
Research subject group fr
Not available
Research subject en
Non disponible
Research subject fr
Non disponible
Application title
Towards a Robust and Trustworthy Machine Learning for Visual Contents in Data-centric Applications
Keyword
Data-centric AI; Privacy preserving Machine Learning as a Service; Robust Learning with noisy labels; Trustworthy Computing; Weakly Supervised Learning
Application summary
Recently, machine learning (ML) performance has drastically increased as a result of powerful computing facilities and large datasets. However, the inherent weaknesses in the reliability and security of data-centric machine learning applications may result in performance degradation, system misbehavior, and/or privacy breach. Building trustworthy data-centric learning systems is among the biggest and most impactful challenges in the next decade. My goal is to enhance the understanding of the vulnerabilities and/or biases caused by imperfect training sets and to develop the fundamental tools, metrics and methods to explain and mitigate them. We seek to formulate the trustworthy machine learning problems and design solutions to ensure the robustness, trustworthiness, and privacy of data-centric machine learning systems. This research program focuses on developing practical approaches to build dependable and trusted models/systems with noisy training data (which contains non-negligible portion of mislabeled samples), manipulated data, biased data, and inappropriate scale of data, etc. One step further, the research outcomes will be used to algorithmically and automatically generate reliable training sets from social media apps, webs search and outsourcing, and/or identify the mislabeled training samples from the third part datasets. In addition, the proposed research will address privacy-preserving image data sharing and publishing. This will allow machine learning applications to efficiently get the "good" data samples in an appropriate scale, therefore yielding reliable models. It is anticipated that the proposed research will result in profound reliability, security, and privacy implications when we deploy ML systems and help to shift the ML paradigms from model-centric to data-centric. The research agenda will attract participation from a diverse and inclusive team of undergraduate and graduate students, postdoctoral fellows, and industry personnel. They will be sought for their unique expertise in AI, multimedia big data analysis, security & privacy, and social networks. The research outcomes will directly benefit the Canadian society and economy by providing small companies (who don't have enough expertise, computing facilities, and data) an opportunity to use AI technologies successfully, by enhancing data quality and model reliability for large companies, and by leveraging the Canadian leadership in artificial intelligence in terms of dependable, high-performance and low-cost features.