Institution profile

University of Ottawa

Academic institutionnorthamerica · ca
Official website
Research library330linked papers
Opportunities0open roles
Selected work

Representative Papers

Advancing machine fault diagnosis: A detailed examination of convolutional neural networks

Dec 19, 2024Measurement science and technology

To address critical challenges in fault diagnosis of complex mechanical systems—including limited generalizability of CNNs, poor adaptability to heterogeneous sensor signals (e.g., vibration, acoustic), and insufficient robustness under dynamic operating conditions—this paper systematically reviews the theoretical evolution and architectural advancements of CNNs in fault diagnosis. It identifies, for the first time, three pivotal technical pathways: data augmentation, transfer learning, and hybrid CNN-RNN/Transformer architectures. Through empirical evaluation across multi-source signals, we delineate CNNs’ performance boundaries, applicability domains, and intrinsic limitations. Furthermore, we establish a comprehensive diagnostic methodology that jointly ensures reliability (high accuracy, strong robustness) and foresight (cross-device generalization, few-shot learning, online adaptation). The work delivers a reusable technical selection guide and an engineering implementation framework, thereby providing both theoretical foundations and practical paradigms for industrial intelligent maintenance.

3 citationsRead paper

A Contextual Online Learning Theory of Brokerage

May 22, 2024arXiv.org

This paper studies the context-aware online bilateral trading problem: a broker must dynamically set transaction prices for privately informed buyers and sellers, leveraging asset- and market-related contextual features, to maximize expected revenue. We establish the first theoretical learning framework for online brokerage with contextual information, distinguishing between two realistic feedback models—full feedback (where both agents’ valuations are revealed) and binary feedback (where only the transaction outcome is observed). Under a bounded-density assumption on valuations, we propose algorithms based on linear contextual modeling and online convex optimization. In the full-feedback setting, our algorithm achieves the optimal regret bound of $O(Ld ln T)$; under binary feedback, it attains $O(sqrt{LdT ln T})$ regret, and we prove a matching lower bound of $Omega(sqrt{LdT})$. Furthermore, we show that the problem becomes statistically unlearnable without the bounded-density condition.

2 citationsRead paper

Development And Testing Of Novel Soft Sleeve Actuators

Nov 08, 2025Social Science Research Network

Conventional rigid wearable assistive devices for aging populations and individuals with neuromusculoskeletal disorders suffer from low force transmission efficiency and poor anatomical conformity. Method: This study proposes an integrated compliant pneumatic soft sleeve actuator fabricated from thermoplastic elastomer via an enhanced fused deposition modeling process, enabling hermetic, elastic actuators capable of linear, bending, torsional, and omnidirectional compound motions—without requiring complex anchoring mechanisms and operating stably at low pneumatic pressure. Contribution/Results: Experimental evaluation demonstrates substantial improvements in force transmission efficiency and wearer comfort, alongside advantages in lightweight design, high integration density, and multi-degree-of-freedom actuation capability. The architecture establishes an engineering-feasible paradigm for next-generation soft wearable assistive systems.

1 citationsRead paper

An Efficient Near-Optimal Algorithm for Adversarial $m$-Set Bandits

Aug 12, 2026

This work addresses the exponential blow-up in action space inherent in adversarial combinatorial multi-armed bandits, where at each round an agent selects $m$ items out of $d$ and observes only an aggregated loss. The paper proposes an efficient algorithm that exploits the structural assumption that the loss is determined by a $d$-dimensional item loss vector, thereby avoiding explicit enumeration of all $\binom{d}{m}$ actions. By introducing a dual representation and parameterizing a low-dimensional sampling distribution, the method integrates online learning with combinatorial optimization to achieve, for the first time, a high-probability regret bound of $O(\sqrt{dT \log(K/\delta)})$ with probability at least $1-\delta$ (where $K = \binom{d}{m}$) in polynomial time. This matches the theoretical performance of EXP3-KW while eliminating exponential space complexity, resolving an open problem posed by Maiti et al.

0 citationsRead paper

Characterizing Peace Through Scientific Keywords

Aug 11, 2026

This study addresses the conceptual heterogeneity of peace, demonstrating that it is not a neutral or unified scientific construct but rather differentially framed across disciplines and geopolitical contexts. By systematically analyzing keywords in global academic literature linked to Sustainable Development Goal 16 (SDG 16), the research integrates disciplinary classifications with national Peace Index scores through associative modeling. Findings reveal that high-peace countries embed violence within frameworks of welfare, health, and social systems, whereas low-peace countries emphasize conflict, security, and geopolitics. Moreover, the natural sciences predominantly adopt technical and forensic perspectives on peace, while the social sciences foreground institutional and legal dimensions. Moving beyond conventional singular conceptions of violence, this work reconceptualizes peace as a multidimensional construct spanning institutional, societal, and technological domains, thereby elucidating the mechanisms underlying its semantic diversification.

0 citationsRead paper
Recent publications

Latest Papers

An Efficient Near-Optimal Algorithm for Adversarial $m$-Set Bandits

Aug 12, 2026

This work addresses the exponential blow-up in action space inherent in adversarial combinatorial multi-armed bandits, where at each round an agent selects $m$ items out of $d$ and observes only an aggregated loss. The paper proposes an efficient algorithm that exploits the structural assumption that the loss is determined by a $d$-dimensional item loss vector, thereby avoiding explicit enumeration of all $\binom{d}{m}$ actions. By introducing a dual representation and parameterizing a low-dimensional sampling distribution, the method integrates online learning with combinatorial optimization to achieve, for the first time, a high-probability regret bound of $O(\sqrt{dT \log(K/\delta)})$ with probability at least $1-\delta$ (where $K = \binom{d}{m}$) in polynomial time. This matches the theoretical performance of EXP3-KW while eliminating exponential space complexity, resolving an open problem posed by Maiti et al.

0 citationsRead paper

Characterizing Peace Through Scientific Keywords

Aug 11, 2026

This study addresses the conceptual heterogeneity of peace, demonstrating that it is not a neutral or unified scientific construct but rather differentially framed across disciplines and geopolitical contexts. By systematically analyzing keywords in global academic literature linked to Sustainable Development Goal 16 (SDG 16), the research integrates disciplinary classifications with national Peace Index scores through associative modeling. Findings reveal that high-peace countries embed violence within frameworks of welfare, health, and social systems, whereas low-peace countries emphasize conflict, security, and geopolitics. Moreover, the natural sciences predominantly adopt technical and forensic perspectives on peace, while the social sciences foreground institutional and legal dimensions. Moving beyond conventional singular conceptions of violence, this work reconceptualizes peace as a multidimensional construct spanning institutional, societal, and technological domains, thereby elucidating the mechanisms underlying its semantic diversification.

0 citationsRead paper

Patterns of Research Funding Across Research Subjects: The Case of NSERC

Aug 11, 2026

This study addresses how research funding allocation may inadvertently reinforce cumulative advantage and exacerbate inter-disciplinary resource inequality. Drawing on longitudinal data from the Natural Sciences and Engineering Research Council of Canada (NSERC), the analysis integrates disciplinary classifications with time-series trends in funding amounts and application volumes. It reveals, for the first time systematically, a sustained decline in the share of funding allocated to pure sciences and mathematics, alongside a pronounced increase for social sciences and medical sciences. The research identifies structural divergence between high-growth and declining disciplines, offering empirical evidence to inform more equitable and forward-looking research funding policies.

0 citationsRead paper

Policy Convergence and Divergence Across National and Within Regional AI Strategies: A Policy Design Element Analysis

Aug 11, 2026

This study addresses the lack of systematic comparative analysis of global artificial intelligence strategies at the level of policy design elements, which has hindered understanding of convergence or divergence both across nations (horizontally) and between regions and their member states (vertically). Combining qualitative content analysis with latent variable induction, the authors construct a structured framework encompassing goals, methods, and principles to code and compare 74 national and 3 regional AI strategies. Findings reveal strong international consensus on economic competitiveness, scientific research support, and ethical AI use, yet persistent divergence regarding human rights, participatory governance, and human-centered principles. The African Union exhibits the highest vertical coherence, the European Union aligns regulatory and economic objectives but diverges on values, and the Nordic–Baltic region displays a hybrid pattern, collectively illuminating the dynamic interplay between norm formation and regional variation in global AI governance.

0 citationsRead paper

On the Spanning Ratio of the Greedy Triangulation for Convex Point Sets

Aug 10, 2026

This study addresses the problem of the excessively large known upper bound on the stretch factor of greedy triangulations for point sets in convex position. By analyzing path lengths between arbitrary pairs of points and leveraging structural properties of greedy triangulations together with techniques from geometric graph theory, the authors significantly improve the theoretical guarantee, reducing the previously known upper bound from approximately 11,739.1 to less than 17.814. Consequently, they rigorously establish that the greedy triangulation of any convex point set is an 18-spanner. This work presents the first tight constant upper bound for this setting, substantially enhancing both the theoretical precision and practical relevance of greedy triangulations in computational geometry.

0 citationsRead paper