Engineering Commons^™

Computational Prediction Of Host-Pathogen Protein Interactions Between Melioidosis Pathogen Burkholderia Pseudomallei And Human, Chathumadavi Ediriweera

Student Research Symposium

Background:

Burkholderia pseudomallei (Bp) is a critical biothreat agent, causes Melioidosis: a disease associated with high case-fatality rates in animals and humans; even with treatment, its mortaility is up to 50%. It also infects plants. Yet few effector proteins have been functionally characterized to date and little work has been carried out to understand the host-pathogen protein-protein interactions (PPI) which are important to develop efficient prevention measures. To address these issues, we deployed diverse in-silico methodologies to identify virulence factors and their interactions with host proteins to characterize bacterial infection mechanisms.

Fair Resource Allocation In An Mec-Enabled Ultra-Dense Iot Network With Noma, Qun Wang

Student Research Symposium

Ultra-dense Internet of Things (IoT) network has greatly facilitated the development of smart environments and the realization of diverse sophisticated applications. Power constrained and resource-limited IoT devices often need to perform computation-intensive and delay-sensitive tasks in such a network. Mobile edge computing and non-orthogonal multiple access are two promising techniques to address the corresponding challenges. In order to improve the fairness and resource efficiency among IoT users, resource allocation problems are formulated in ultra-dense MEC-enabled IoT networks with NOMA considered. An iterative algorithm based on successive convex approximation technique is proposed to solve those challenging non-convex problems.

Exploiting Wall Street: An Algorithmic Approach To Investing, Sterling Lemon

Student Research Symposium

Technology has advanced dramatically over the years, yet the old investment strategy of "buy and hold" has remained prevalent. While this strategy performed well historically, it serves as a stark contrast to modern investing which, by utilizing machine learning and advanced data analytics, is generating unprecedented returns This research project will demonstrate the strength behind algorithmic investing and how the investment strategy of "buy and hold" will become a thing of the past.

Propulsion Unit Optimization For Small Uas, Cory Goates

Student Research Symposium

Selection of a propulsion system for unmanned aerial vehicles (UAV) is a complex and iterative process. Attempts have been made to provide effective modelling of propulsion systems. However, due to the discrete nature of the design space, effective optimization tools have yet to be developed. We present an optimizer which presents designers with feasible propulsion systems based on given parameters. This optimizer also allows designers to view trend in the design space and more quickly determine an optimum solution.

Augmenting Anaerobic Digestion Of Microalgal Biomass, Anna Doloman

Student Research Symposium

Anaerobic digestion of microalgal biomass cannot be achieved without specialized hydrolytic microorganisms. Potentially algalytic bacteria belonging to Citrobacter and Alcaligenes species were isolated from a wastewater lagoon system. A combination of two potentially algalytic bacteria was successfully incorporated into the granular anaerobic consortia. A series of anaerobic cultures were prepared with different microbial combinations to test the methane production from algal biomass collected from a local wastewater treating trickling filter. The anaerobic microbial community mixed with two algalytic bacteria produced 10% more methane when compared to the methane produced by a native granular consortium. The presence of the algalytic bacteria …

Robust Structured Group Local Sparse Tracker Using Convolutional Neural Network Features, Mohammadreza Javanmardi

Student Research Symposium

Sparse representation has recently been successfully applied in visual tracking. It utilizes a set of templates to represent target candidates and find the best one with the minimum reconstruction error as the tracking result. In this presentation, we propose a robust deep features-based structured group local sparse tracker, which exploits the convolutional neural network (CNN) features of local patches inside target candidates and represents them by a set of templates in the particle filter framework. To extract the local CNN features, we first set the size of each target candidate to 64 by 64 pixels to contain sufficient object-level information …