Travel Time Prediction using Deep Learning to solve TDVRPTWo Developed a deep learning algorithm that outperforms existing ones in predicting travel time by 2.9%.o Applied Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Temporal Convolutional Networks (TCN) to learn spatiotemporal information from GPS large data sets (1.1 billion GPS points).o Utilized the estimated travel time to optimize distribution planning by solving time-dependent vehicle routing problem (VRP) in the supply chain.o Implemented and applied Genetic Algorithms (GA), Local Search (LS), and Large Neighborhood Search (LNS) techniques to solve Time-dependent VRP in Python.o Presented research findings and results at two conferences in Canada.o Demonstrated expertise in deep learning techniques, route optimization, spatiotemporal data analysis, and supply chain management.

Developed familiarity with the intelligent transportation system and GIS, and analyzed existing dashboards related to road transportation data (e.g., license plate data).

o Collaborated with a team of 12 B.Sc., M.Sc., and Ph.D. students to develop an app that translates sound or text data to Persian sign language using NLP and deep learning for the deaf communities in Iran. o Worked closely with stakeholders (deaf communities and translators) in understanding key business needs.o Gathered and analyzed customer requirements to develop recommendations about the necessary features of the app.