Stats 253: Analysis of Spatial and Temporal Data

Goal

The goal of the project is to produce something related to spatial statistics or time series that others can use, i.e., not something that you throw away after the quarter is over! For example, you might publish a short conference paper or release open source code. This is ambitious, but not as difficult as it sounds! You will find that the methods taught in this class have not yet made inroads into many fields that work with spatial and temporal data. You can make a valuable contribution by applying these methods to answer questions in those fields.

Example Ideas

Digital humanities: A recent trend in history, literature, and art is to analyze the statistics of large digital databases of maps, books, and paintings. Find a recent paper in this field, and re-analyze the data using methods from this class. Are you able to obtain any new insights?
Audio / music processing: The NMF model described in Week 5 enables many audio/music applications, from automatic transcription to genre classification. Try applying NMF to solve an audio/music problem that interests you. (This is one of my research areas, so please talk to me if you're interested.)
Your own research: If you are involved in a research project, feel free to use this opportunity to try out methods on a dataset you are already working on.
I will also mention specific project ideas as they come up in lecture.

Requirements

The project is due August 18 at noon. No extensions are possible because I have to submit grades. The requirements are determined by your personal goals for the project. For example, if you are writing a paper, you are encouraged to identify a publication venue and then to submit a paper satisfying the requirements of that venue.

Timeline

E-mail us a project proposal: July 24.
Final draft due: August 14.