Senior Project Topics

SPRING 2025 Senior Project Descriptions

Professor Erik Grimmelmann (Section L)

This course provides a theoretical and hands-on introduction to the modeling and analysis of complex systems.  Both analytical and stochastic approaches to modeling will be considered. Among the topics covered will be: brief reviews of the relevant topics in math and numerical analysis, discrete time models, continuous time models, bifurcations, chaos theory.

The modeling of a number of complex systems will be examined, including: The Game of Life, Turing patterns in animal skins, Covid-19 epidemiology, Forest fire propagation, Stock and option pricing, Vehicular traffic jams, Housing segregation, Protein synthesis.

This course is the first in a two-course sequence.  In the following semester, students enroll in CSc 59867 – Senior Project II – in which they will implement small group projects related to the modeling and analysis of complex systems and/or to artificial intelligence or machine learning.

Instructor: Yunhua Zhao (Section E)

Deep Learning concerns multi-level data representation with every level providing hierarchical explanations of data.

The course will equip the students with a general and practice-oriented foundation of deep learning, including the underlying theory, the range of applications, and learning from very large data sets. To this aim, the students will also be introduced to computer vision, natural language processing and modern advanced algorithms and systems, which also serves as a foundation for the Senior Project proposals and projects of the following semester.

The senior project will target designing a deep neural network system which solves a specific problem. The students will develop deep neural networks which is relevant for the aimed project goal using TensorFlow/Pytorch on a large dataset. The performance of the algorithm will be demonstrated on different validation metrics, and others suggested by the students.

Instructor: Ovsanna Bogosyan Estrada (Section 5BC)

Control of Autonomous Vehicles and Mobile Platforms: Many applications and services are depending more and more on autonomous vehicles (AVs) for different applications, from transportation, factory automation, and urban logistics, to smart farming/agriculture and disaster management to name a few. The learned material in this course will not only prepare the CS & CpE students for further research and technology development in this evolving area, but will also help understand other automated systems and mobile platforms.

 The first part of this course will equip the students with  a general and practice-oriented  foundation in autonomous vehicle (AV) control, focusing on the three layers; namely, perception & localization, motion & behavioral planning, and trajectory- tracking control.  To this aim, the students will also be introduced to general modeling and computer-control basics, which would be helpful for a myriad of other computer science and engineering applications. AV control requires the vehicle to perceive its environment, know its location, and based on that information plan a behavior and motion, which should be adequately realized by the tracking layer of the vehicle. This course will introduce the students to algorithms that are most commonly used in the practice of AV control for each layer of autonomy, with practical examples in-class and weekly assignments. The students will also learn about the most popular AV control, simulation, and animation platforms; i.e. ROS2, Gazebo&RViz through lectures, examples, and assignments. There will be an end-of-semester team project that combines all the layers of AV control, which also serves as a foundation for the Senior Projects of the following semester.

 The projects will target  a single  layer of AV control, or may combine multiple or all  layers. The students will develop algorithms relevant for the aimed performance at a given autonomy layer of the vehicle or mobile platform,  and use Phyton, or C languages for the programming. The performance of the algorithm will be demonstrated on an animation platform (Gazebo, Carla, Autoware etc.) and quantitative performance evaluation will be performed based on metrics, such as minimum time, minimum error, number of crushes, and others suggested by the students.