Student dissertation/thesis/project topics

I describe below some topics on which you can work with me for your MS project or thesis, or your PhD dissertation. I will first mention some common skills that you will require to work on these, and the types of skills you will acquire.

Pre-requisite skills:

• C/C++ programming knowledge: All the topics have a significant programming component. Most of my current software is written in C, and you will need proficiency in it. Alternatively, proficiency in C++ may be sufficient.

• Mathematics background: Many of the topics have some mathematical component to them. Depending on your ability and inclination, you can work more along algorithmic directions, which would require greater mathematical maturity, or more along a software development direction, which would require less mathematical skills. However, some mathematical background will be useful in either case. I will later mention specific requirements along with each topic. An MS project is primarily programming oriented, and less mathematical background may be sufficient for it.

• A yearning for learning: You should be interested in maximizing your learning, rather than in minimizing your effort, especially for a PhD dissertation and MS thesis! You should also be interested in doing good quality work.

Skills that you will acquire:

• Software development: You will get experience in writing a significant piece of software. Furthermore, some of these software will probably have users outside FSU, and so you will get exposure to user support, creating a user-friendly interface, etc.

  The other skills listed below are particularly applicable to PhD students, and to a lesser extent to MS Thesis students.

• Algorithms: Most of the PhD dissertation topics have a significant algorithmic component to them, and you will also learn some new mathematical techniques.

• Research skills: There is a creative aspect to research that is hard to teach in a classroom. Working on these research topics will expose you to real research. You will also read journal articles, interact with other researchers, etc.

• Written communication: I will ask you to submit short reports of your current activities regularly, in order to give you practice in professional writing. You will also probably write papers for conference and journal publications, apart from your thesis or dissertation.

• Oral presentations: I will ask you to give frequent oral presentations, and perhaps occasionally have others too attend and give suggestions, so that you develop confidence in your oral presentation skills.

Miscellaneous matter

• Time frame: I expect that a project will take at least a semester, and a MS thesis will take around a year. However, you should aim to finish your thesis about half a semester before you graduate, in order to give yourself enough time for formalities. Of course, this is just an estimate, and assumes that you put in sufficient effort of sufficient quality. The time required for a PhD dissertation is harder to estimate. But it should normally take at least three years of research effort, after completing most of your course work.

• Difference between a thesis and a project: A project is more oriented toward developing a good piece of software, whereas a thesis involves developing ideas too. A thesis will give you an opportunity for developing other skills, such as research and presentations at conferences. On the other hand, a thesis will take more time, but may help you better in your job search.

• Research assistantships: I do have some funds at this time and am awaiting results on other grants. However, I am not likely to provide an asistantship without evaluating students for a semester at least, at this time.

Topics

1. Scalable algorithms, with applications in nano-materials

   • Level: PhD dissertation, MS thesis, or MS project.

   • Description: Massive parallelism, with hundreds of thousands of processors, promises abundant computational power. Furthermore, scientific applications needing such power are rapidly emerging. This combination promises a leap in scientific understanding, leading to useful products, such as fuel efficient cars, disaster resistant structures, and new methods of treating diseases. An important impediment to such developments is the difficulty in using massively parallel machines effectively -- typically, few applications obtain good efficiency on even a thousand processors. The aim of our research is to develop parallelization strategies which will enable applications to scale efficiently on much larger numbers of processors than they currently do. In particular, we are working on optimizing assignment of tasks to nodes, dynamic load balancing, and efficient GPU implementations of some Quantum Monte Carlo codes. This project is in collaboration with a researcher at Oak Ridge National lab.

   • Mathematics pre-requisites: PhD thesis: Familiarity with numerical linear algebra and common numerical methods will be useful, as will interest in learning combinatorial optimization.

2. Tools for Heterogeneous Multi-Core Processors

   • Level: PhD dissertation, MS thesis, or MS project.

   • Description: Multi-core processors, as the name implies, have multiple cores on a chip. In heterogeneous multi-core processors, these cores are not identical. An example is a conventional procerssor accelerated by a GPU. They can provide great computational power, but are difficult to program if one wishes to make effective use of the hardware. One of the important problems on which we are working is the development of tools for optimzation of workflows for GPU-accelerated computation. This is in collaboration with a researcher at Infosys SETLabs.

   • Mathematics pre-requisites: None. An interest in learning optimization techniques will be useful.

3. Supercomputing for Policy Decisions

   • Level: PhD dissertation, MS thesis, or MS project.

   • Description: Computer simulations can be an effective tool for policy makers to understand the consequences of decsions that they make. For example, policy makers may wish to mitigate the more serious consequences of climate change. The consequences depend not just on the expected climate change, but also on the response of humans to this change and to policy decisions taken. Societal change can, in turn, influence climate change in the long term. There is, therefore, a need for deeper understanding of the complex interactions between climate change, human response, and policy decisions, and also an ability to manipulate this complexity. The goal of this project is to develop models, computational techniques, and tools that will provide such understanding in the context of taking policy decisions. This project is in collaboration with a researcher at the Decision Theater at Arizona State University and others. We will focus on the computational techniques.

   • Mathematics pre-requisites: No specific area is absolutely required. However, for a PhD, different areas of mathematical expertise, such as optimzation and data analysis, could be useful, depending on the direction that you wish to pursue.

4. Dynamic data-driven applications

   • Level: MS thesis.

   • Description: Computer simulations and data from experiments can play complementary roles, with the simulations attempting to identify the underlying mechanisms that the physical system in the experiment is undergoing, and dynamically obtained data from the experiments correcting predictions from the simulations. This is a new topic on which we have formulated our ideas, but have not started work. It is currently offered as an MS Thesis, but can be extended in the future to the PhD level.

   • Mathematics pre-requisites: PhD thesis: Familiarity with numerical methods will be useful. Interest in learning about data analysis is desirable.

5. Algorithms for Load Balancing

   • Level: MS thesis or MS project.

   • Description: In a parallel computation, one needs to ensure that each processor has roughly equal amount of work to do. This work involves some new ideas I have regarding algorithms to perform such load balancing.

   • Mathematics pre-requisites: Good knowledge of linear algebra, including eigenvectors, will be needed.