Undergraduate Research Projects - University of Houston
Skip to main content

Undergraduate Research Projects

The following are examples of recent or current undergraduate research projects for the courses COSC 3396 & COSC 4396.

 


Cybersecurity

Advisor: Dr. Stephen Huang

Identifying IntrudersUsing Long Connection Chains

Project Description

In a long connection chain from a hacker to a server, packets going between the client and the server will cross each other from time to time. The number of packet cross-over is proportion to the length of the chain. We cannot measure the cross-over at the server, but we are able to see some side-effects of packet cross-over. Students will study the data collected from experiment and design detection algorithms.

Identifying Intruders by Analyzing their Cyber Behavior

Project Description

The objective of this study is to find all attributes that can separate a hacker from a normal user in terms of their cyber behavior. Students will study the data collected from normal users and build a profile of normal cyber behavior.


Economics of Security

Advisor: Dr. Aron Laszka, alaszka@uh.edu, aronlaszka.com

Bug Bounty Programs

Project Description

The project will collect and analyze real-world data, model the process of vulnerability discovery and the incentives of sponsoring organization and hackers, and propose novel approaches for improving the efficiency of bug bounty programs.


Email Security

Advisor: Dr. Rakesh Verma, http://www2.cs.uh.edu/~rmverma/

Phishing Email Study

Project Description

The project has a number of research directions, including: user awareness and education, automatic detection, etc.

Additional Comments

Dr. Verma is especially interested in the intersection of data analytics and security. He also works on formal methods, design and verification of protocols and natural language processing.


Natural Language Processing (NLP)

Advisor: Thamar Solorio, thamar.solorio@gmail.com, ritual.uh.edu

NLP for Social Media Data

Project Description

Information Extraction in Social Medial Data


Real-Time Systems,Cyber-Physical Systems

Advisor: Prof. Albert M. K. Cheng, cheng@cs.uh.edu

Building Safe, Fault-Tolerant, and Secure Real-Time, Embedded, Cyber-Physical Systems; Real-time Machine Learning

Project Description

The focus is on: (1) controller software implemented as functional reactive programs (FRP), (2) secure embedded systems, (3) real-time virtualization/cloud computing, (4) scheduling, (5) real-time machine learning and data analytics, and (6) FRP compiler implementation. You will have access to real-time systems specification, scheduling, development (Wind River WorkBench/Tornado), and verification tools, Lego Mindstorms
robotic kits, an Intel XScale development board with current and
temperature sensors, and quad-core Intel Core i3 and i7 workstations with
8/16 GB RAM, 1 & 3 TB disks, and 24/27-inch monitors.

I am looking for highly motivated undergraduates with good programming
skills (C, C++, Unix/Linux), strong mathematics/logic background
(calculus, differential equations, linear algebra, probability,
first-order logic), modeling tools for mechanical systems, and interest inembedded and real-time systems (scheduling, languages, verification).

Additional Comments

U.S. citizen or permanent resident, Recommended GPA >= 3.5/4.0, Computer science, electrical engineering, mechanical engineering, or mathematics undergraduate (preferably Junior and Senior-level, but Sophomore-level will also be considered)



Security & AI/Machine Learning

Advisor: Dr. Aron Laszka, alaszka@uh.edu, aronlaszka.com

Security alert prioritization using deep reinforcement learning

Project Description

This project combines intrusion detection systems, game theory, and AI/ML to create a solution that will help system operators to prioritize the investigation of security alert.


Security & Software Engineering

Advisor: Dr. Aron Laszka, alaszka@uh.edu, aronlaszka.com

Correct-By-Construction Smart Contracts

Project Description

With the advent of blockchain-based distributed computing platforms, such as Ethereum, smart contracts have seen wide adoption. However, many of the contracts deployed in practice suffer from vulnerabilities, which presents a significant risk (e.g., in the infamous "The DAO" attack, $50 million worth of cryptocurrency was stolen). This project will develop both theory and practical tools for facilitating the construction of secure smart contracts.


Security of Cyber-Physical Systems

Advisor: Dr. Aron Laszka, alaszka@uh.edu, aronlaszka.com

Blockchain-Based Decentralized Energy Trading Platform

Project Description

The project will develop a privacy-preserving decentralized trading platform for transactive energy, building on blockchain-based computing platforms, smart contracts, cyber-physical energy system models, and applied cryptography.


Software Testing & Analysis

Advisor: Dr. Alipour

Testing Web pages

Project Description

The goals of testing web appsare to find dead links and incorrect rendering of pages. We are particularly interested in (1) techniques that efficiently explore and evaluate the functionality of web apps using the usage patterns, (2) techniques to compare the functionality of mobile version and PC version of web apps and determine the differences.

Additional Comments

Requirement: B or better in COSC 3320


Testing Infrastructure for C & Python Programming Languages

Project Description

We build tools to automatically change the statements in C or Python programs and compare the behavior of the original program with the modified program.

Additional Comments

Requirement: B or better in COSC 3320


Network Analysis in GitHub and LinkedIn

Project Description

Given the data in GitHub and Stack Overflow, we'd like to annotate the GitHub profiles with the links to users’ profiles. We are also interested in analyzing the GitHub social network and find the users/organizations that were pivotal in popularity of different technologies.

Additional Comments

Requirement: B or better in COSC 3320



Video Analytics

Advisor: Shishir Shah

Video Event Detection

Project Description

Analysis of events in public spaces observed by surveillance cameras is critical in understanding how video surveillance can become an effective resource.

In this project, we are interested in exploring and enhancing behavioral analytics by examining low- and mid-level trends of common observed events in surveillance cameras. In doing so, this project will emphasize both algorithmic and system design elements to develop and demonstrate innovative video analytic solutions.

Additional Comments

This project can have multiple sub-projects and the specific topic(s) can be finalized based on discussions with interested students.