Innovative Projects Realized

Portfolio management by reinforcement learning

This project is addressing the problem of portfolio optimization by using reinforcement learning, an area of machine learning that has recently attracted many researchers. Its advantages compared to the conventional models of portfolio optimization are coming from its ability in incorporating many features of the assets into the asset allocation problem without relying on the predicted returns outputted from another model. This model is able of considering risk measures and providing asset allocation strategies that are in line with the risk preferences of the investors. It also considers exogenous market constraints like the transaction costs, which are very important factors in the performance of the suggested strategies of the model.

Optimization of a Stochastic Heterogeneous Vehicle Routing Problem for the Trucking Industry

In the transportation industry customer satisfaction, driver retention and profitability are some of the most critical factors in company success. As regulatory environments change, and competition grows Contrans Corp., a flatbed freight company, is seeking to develop a user-friendly application that allows team members to rapidly and accurately determine an optimal dispatch schedule for the fleet. This tool takes in information, such as road conditions, order types, vehicle specifications, driver specifications and time windows and solves a vehicle routing problem to maximize company profits and driver satisfaction, all whilst ensuring customer demand is met. This tool will be implemented in a graphic user interface and will serve as a case study for how operations research can ensure sustainability in the trucking industry.

Spatial and temporal trends in nesting habitat use and availability for cavity nesting waterfowl in the lower Saint John River floodplain

Cavity-nesting ducks, including the wood duck, common goldeneye, and hooded merganser, are of interest in wildlife management programs due to their value to hunters and conservation groups. The lower Saint John River floodplain (New Brunswick, Canada) is a major breeding region for these species in Atlantic Canada that has experienced significant changes in recent decades. These ducks depend on natural cavities that form in trees to nest, but they will also use nest boxes when available. To improve the availability of suitable nesting space, Ducks Unlimited Canada (DUC) runs a nest box program in the lower Saint John River region to help maintain and increase their populations. An assessment of the program, including how use and availability of nest boxes compares to use and availability of natural cavities, is being undertaken in order to better understand the impacts of the nest box program on these species.

Programme Tremplin Santé – Power Up Program

Many existing programs encourage young people to move and eat better in order to counteract overweight and obesity trends and reduce chronic diseases. Despite this, rates remain high. This proposal is an evaluation project of Shape Up program which provides training, support and tools for summer camp workers to have a lasting impact on young people (from 8 to 11 years old) to move more and eat better on healthy habits. This project aims to evaluate the impact of Shape Up on counsellors and campers with respect to healthy eating and physical activity. It also aims to evaluate the capacity of the camps adhering to Shape Up to promote healthy eating and physical activity environments. The study follows a design that conducts measures at baseline followed by immediate post and 1-year post after participants have been at the camp.

A study of macrophage interactions with polymeric drug carriers

The focus of this study is on the understanding of the cellular interactions with polymeric drug carrier for intraarticular injection of therapeutic. The degradability of the Eupraxia’s polymeric drug carrier will be studied in contact with macrophages and/or enzymes. At each time point in the cell study the polymer surface will be characterized by SEM, and the enzymes and ROS production will be measured to compare to the control groups. The enzymatic degradation of the polymeric drug carriers will be analyzed by measuring the mass loss, Tg, surface chemistry and morphology analysis at each time point. The structure and molecular weight of the degradation products will be characterized by using NMR and MS techniques. The potential degradation mechanism will be proposed for the polymers.

Synthesis of CO2-Based Polymers for Carbon Capture

Carbon dioxide is produced from the combustion of fossil fuels and as a by-product of many chemical processes. The increasing concentration of CO2 in the atmosphere has now been decisively linked to global climate change. The use CO2 as a carbon-building block has attracted much attention due to its low cost, ease availability, and its potential to substitute fossil fuel-based feedstocks, in addition to the clear positive environmental effect of removing it from the atmosphere. In this project, we plan to make capture CO2 and polymerize it with epoxides and other derivatives to produce cyclic carbonates or functionalized polycarbonates. Polycarbonates produced by this process will have applications as ceramic binders, adhesives, coatings, and packaging materials, as well as in the synthesis of engineering thermoplastics. The synthesis of more diverse CO2-based copolymers is an important and challenging research and development topic, with clear positive economic and environmental applications. The functionalization of this class of polymers can further leverage the application of CO2 as a feedstock for a wider range of applications.

Equivariant Siamese Neural Networks

The world we live in is ripe with symmetry. From the bilateral symmetry we see in humans to the symmetries which are used to describe fundamental particles in physics. Most modern machine learning methods however do not have an inherent modeling of symmetry in them. By developing algorithms which do have an explicit modeling of symmetry we can decrease the amount we need to teach these algorithms, making them much cheaper to create. We propose a network that can compare images in such a way that it is not affected by changing the orientation of objects in the image. This is useful for things such as facial recognition, where it could be used to verify a face regardless of its orientation. The purpose for this project is to create a method for robust object comparison so that it may be used for industrial applications.

Continue the development of PanSharp software system

The existing PanSharp software developed at University of New Brunswick will be extended

to make it equivalent to commercially available software. An image viewer will be developed

to view the results of PanSharp software, At present. the PanSharp software only supports

GeoTIFF image file format. Therefore, support for other commonly used image file formats

such as ERDAS, ENVI, and ESRI will be developed. A web-based automatic licensing

schema will developed to protect the PanSharp software, To make the software easy to use,

a complete help file and tutorial, guiding the processing steps of the PanSharp software, will

be developed, Finally, the software will be tested for its functionality and quality. The

problems found at the time of testing will be fixed.

Performance Analysis of a Network of Dual-Lane Roundabouts – Part 2

Modern roundabouts have become a subject of great interest and attention over the last few years considering their potential to reduce vehicular delay & emissions and increase safety. The objectives of this study to examine the short- and long-term benefits performance of roundabouts from a multi-modal perspective. Using real observed data from vehicular traffic, pedestrians, cyclists and transit vehicles and complementing it with microsimulation data, this research will assess the delay, environmental impacts and safety of a network of three roundabouts. By conducting this research, the Intern would have the opportunity to incorporate research expertise in a real-life engineering project and benefit from the practical engineering viewpoints at the partner organization. Industrial partner, ISL Engineering and Land Services, is seeking for a comprehensive and comparative analysis among the performance of roundabouts and a network of signalized intersections from a multi-modal perspective. Such analysis provides valuable recommendations for future applications both design approaches for different traffic conditions.

The Vape Mediascape: Youth, E-cigarettes, and Health Policy

A survey, comprised of 20 questions, and organized according to the following four categories/themes: 1) Patterns of use; 2) Perceived health risks; 3) Access to e-cigarettes; and 4) Exposure to cultural content (i.e. marketing and advertisements) will be distributed to a minimum of 100 students from grades 7-12 in the Ottawa private school system. The survey will also include attached images of e-cigarettes advertisements with open-ended questions to better grasp how survey respondents perceive e-cigarette ads. Ultimately, this research will probe the rhetoric around the use and cultural content of e-cigarettes and how it resonates with a number of youth, comprising of both public school youth and at-risk youth attending school in non-mainstream setting.

Incidence de l’anémie à la sortie de réanimation chez les enfants opérés d’une cardiopathie congénitale chirurgie cardiaque

Malgré des améliorations notables dans les techniques chirurgicales, une grande proportion d’enfants atteints de malformations cardiaque continuent de souffrir d’une déficience neurologique significative affectant leur rendement scolaire et leur intégration sociale. Malgré un intérêt soutenu dans la matière, la physiopathologie des lésions neurologiques dans cette population reste seulement partiellement comprise. L’anémie est une complication fréquente des enfants cardiaques gravement malades et a été associée à des résultats neurologiques indésirables chez les enfants âgés de moins de 2 ans. Près de 50% des patients souffrant d’une détresse vitale sont anémiques à la sortie de réanimation, plus d’un tiers sont anémiques à la sortie de l’hôpital et beaucoup demeurent anémiques jusqu’à 6 mois après. Cependant, il n’y a pas de données spécifiques pour les enfants après une chirurgie cardiaque. Le taux d’incidence du retard neuro-développemental chez les enfants est élevé après la chirurgie cardiaque. Les conséquences potentielles à long terme de l’anémie chronique sur le cerveau en développement des nouveau-nés et des nourrissons opérés du cœur peuvent être importantes. Il est impératif que nous vérifiions si l’anémie s’associe à des lésions neurologiques dans cette population.

ANALYSE DE LA CAPACITÉ D’ADHÉRENCE CELLULAIRE DE FIBROBLASTES TRANSDIFFÉRENCIÉS DANS DES KÉRATINOCYTES CONCERNANT DES ÉCHAFAUDAGES BIOLOGIQUES

Les greffes de peau traditionnelles, tout en respectant la norme absolue en matière de traitement des plaies, ont un succès limité chez les chevaux en raison de la rareté de la peau des donneurs. De ce fait, l’étude des fibroblastes en culture 3D peut aider à comprendre les mécanismes de l’interaction cellule-matrice, car ces cellules sont relativement à l’état quiescent et sont principalement responsables de la production et du renouvellement de la matrice extracellulaire (MEC). molécules. De plus, nous avons cherché à étudier l’interaction cellulaire dans le processus de transdifférenciation des fibroblastes dans les kératinocytes sur des échafaudages biologiques et à comparer les échafaudages de la peau d’un groupe de chevaux vivant dans les régions inondées du Brésil – le cheval Baixadeiro – avec du pur-sang.