Scholarship @ UWindsor

Scholarship @ UWindsor is the institutional repository of the University of Windsor (UWindsor), showcasing and preserving the UWindsor community’s scholarly outputs, as well as items from the Leddy Library’s Archives & Special Collections. Its mission is to disseminate and preserve knowledge created or housed at the University of Windsor.

Contact scholarship@uwindsor.ca for more information.

Communities in Scholarship @ UWindsor

Select a community to browse its collections.

Now showing 1 - 4 of 4

Conferences and Conference Proceedings 3171
Papers, presentations and abstracts of conferences held at the University of Windsor, in person and virtually.
Digitized Local Collections 1552
Digitized local items from the collections of the Leddy Library, University of Windsor, and community partners.
Faculties, Departments and Research Units 4092
Open Access Faculty publications, reports and working papers from academic departments at the University of Windsor.
Theses, Dissertations, and Major Papers 9579
Formal graduate original research from the University of Windsor's Masters and Doctoral programs.

Recent Submissions

Access status: Open Access ,
Water source dynamics influence macroinvertebrate communities across groundwater-fed streams in a glacierized catchment
(Springer Science+Business Media, 2023-03-16) Jill Crossman; Chris Bradley; Fredric M. Windsor; Alexander M. Milner
Access status: Open Access ,
A mechanistic model explaining ligand affinity for, and partial agonism of, cannabinoid receptor 1
(American Chemical Society (ACS), 2023-01-25) Fred Shahbazi; Daniel Meister; Sanam Mohammadzadeh; John F. Trant
CB1, a member of the G protein-coupled receptor class, is the putative protein target of THC, the psychoactive component of cannabis. To better identify new synthetic cannabinoids with increased activity, all cannabinoids with reported experimental binding to the CB1 receptor were modelled in silico to build a predictive model for CB1 affinity of small molecules. Computationally derived affinity is not sufficient in and of itself to predict binding, but coupled with the experimental evidence that ligands enter the receptor from the membrane rather than solvent, we provide a model that accurately describes the binding of these molecules by incorporating a correction factor for relative hydrophobicity. In addition, we propose a mechanism of action for partial CB1 agonists based on molecular dynamics simulations of THC homologues, modelling long time scale structural changes in the CB1 receptor. Together, the affinity model, and the mechanism of agonism/antagonism can allow for the computational prediction of both the effective behaviour and potency of novel cannabinoids, and several such predictions are made.
Access status: Open Access ,
3D Object Detection with Attention: Shell-Based Modeling
(Tech Science Press, 2023-01-01) Xiaorui Zhang; Ziquan Zhao; Wei Sun; Qi Cui
LIDAR point cloud-based 3D object detection aims to sense the surrounding environment by anchoring objects with the Bounding Box (BBox). However, under the three-dimensional space of autonomous driving scenes, the previous object detection methods, due to the pre-processing of the original LIDAR point cloud into voxels or pillars, lose the coordinate information of the original point cloud, slow detection speed, and gain inaccurate bounding box positioning. To address the issues above, this study proposes a new two-stage network structure to extract point cloud features directly by PointNet++, which effectively preserves the original point cloud coordinate information. To improve the detection accuracy, a shell-based modeling method is proposed. It roughly determines which spherical shell the coordinates belong to. Then, the results are refined to ground truth, thereby narrowing the localization range and improving the detection accuracy. To improve the recall of 3D object detection with bounding boxes, this paper designs a self-attention module for 3D object detection with a skip connection structure. Some of these features are highlighted by weighting them on the feature dimensions. After training, it makes the feature weights that are favorable for object detection get larger. Thus, the extracted features are more adapted to the object detection task. Extensive comparison experiments and ablation experiments conducted on the KITTI dataset verify the effectiveness of our proposed method in improving recall and precision.
Access status: Open Access ,
Deep Clustering with a Dynamic Autoencoder: From Reconstruction towards Centroids Construction
(Ryerson University Library and Archives, 2023-05-03) Nairouz Mrabah; Naimul Khan; Riadh Ksantini; Zied Lachiri
<p>In unsupervised learning, there is no apparent straightforward cost function that can capture the significant factors of variations and similarities. Since natural systems have smooth dynamics, an opportunity is lost if an unsupervised objective function remains static during the training process. The absence of concrete supervision suggests that smooth dynamics should be integrated. Compared to classical static cost functions, dynamic objective functions allow to better make use of the gradual and uncertain knowledge acquired through pseudo-supervision. In this paper, we propose Dynamic Autoencoder (DynAE), a novel model for deep clustering that overcomes a clustering-reconstruction trade-off, by gradually and smoothly eliminating the reconstruction objective function in favor of a construction one. Experimental evaluations on benchmark datasets show that our approach achieves state-of-the-art results compared to the most relevant deep clustering methods.</p>
Access status: Open Access ,
A SAT Solver + Computer Algebra Attack on the Minimum Kochen-Specker Problem
(Springer Science and Business Media LLC, 2023-06-21) Zhengyu Li; Curtis Bright; Vijay Ganesh
One of the most fundamental results in the foundations of quantum mechanics is the Kochen--Specker (KS) theorem, a `no-go' theorem which states that contextuality is an essential feature of any hidden-variable theory. The theorem hinges on the existence of a mathematical object called a KS vector system. Although the existence of a KS vector system was first established by Kochen and Specker, the problem of the minimum size of such a system has stubbornly remained open for over 50 years. In this paper, we present a new method based on a combination of a SAT solver and a computer algebra system (CAS) to address this problem. We improve the lower bound on the minimum number of vectors in a KS system from 22 to 23 and improve the efficiency of the search by a factor of over 1000 when compared to the most recent computational methods. Finding a minimum KS system would simplify experimental tests of the KS theorem and have direct applications in quantum information processing, specifically in the security of quantum cryptographic protocols based on complementarity, zero-error classical communication, and dimension witnessing.