Ryan Dempsey Email and Phone Number

• Research and develop machine learning-enabled propagation models

• Optimized convolutional neural network to classify trees with 87% accuracy on geographically independent test folds in Ottawa• Developed machine learning module to label clutter maps from satellite images via Google Maps API requests, for use in propagation models• Developed machine learning modules to model Long-Term Evolution Physical Resource Block Utilization using Tensorflow 2 in Python• Implemented HTTP-based API for visualization of Geospatial Physical Resource Block Utilization• Contributed to R-based module that analyzed existing Wireless Broadband Service deployments in order to support sector with decisions on new license assignments• Estimated licensees’ achievable download speeds in current frequency band and mapped to required bandwidths in new band• Modelled rural Wireless Broadband Services using Longley-Rice propagation model• Assessed technical capabilities of Universal Broadband Fund applicants’ broadband networks• Compared Universal Broadband Fund applicant projects for cost, longevity, and impact• Aggregated crowd-sourced cellular deployment data by population centre using BigQuery• Estimated population centre statistics from Statistics Canada dissemination area data using geospatial tools in R and BigQuery• Developed statistical models in R to quantify the Canadian digital divide based on cellular deployment as a function of estimated socioeconomic factors across population centres• Adapted Temporal Fusion Transformer artificial neural network architecture to predict hourly LMR spectrum occupancy data using various Python packages and object-oriented programming techniques• Optimized machine learning model to achieve 93% median channel accuracy over 1-week prediction of 900 channels

• Developed solutions to example problems on board in tutorials/office hours• Contrasted counter-examples and exceptions with students to ensure meaningful comprehension• Interpreted and evaluated written work to clarify confusion and encourage improvement

• Developed application framework to stabilize video, track points, and analyze their trajectories, using feedback from users to implement new features through MATLAB• Re-engineered existing tracking and plotting software by investigating source code and features to create user-friendly and functional programs for use in the lab• Tested and updated software using MATLAB’s debugger to investigate variables and locate and resolve bugs, to maximize ease of use• Consulted with professor to generate concise visual representations of displacement data

• Calculated cash register deficit/excess and organized and reported sales & inventory by accurately counting funds and food inventory, in order to isolate financial discrepancies• Maintained store security and sanitation by cleaning and closing the store in order to keep the store in optimal shape• Accounted for job distribution of newer employees by delegating tasks to staff based on various skillsets, in order to help new staff learn, while maintaining store efficiency• Maximized guest satisfaction by recalling regular customer orders, and providing deals when appropriate