GRAD 2006 Poster Winners
March 2006
Wednesday, March 15, 2006 saw WPI's first annual Graduate Research Achievement Day (GRAD2006) poster symposium. Over 143 posters presenting research in the areas of engineering, science, and the life sciences were on display! An awards ceremony honored the top three posters in each category.
The CS department, included in the science area, had more posters than any other department!
In the area of science we won two prizes. Congratulations go to John Hayward for his 1st place poster "Mining Oncology Data: Knowledge Discovery in Clinical Performance of Cancer Patients" and to his advisor, Prof. Carolina Ruiz. Congratulations too go to Fan Wu for his 2nd place poster "Wavelet-Based Multiresolution for Mobile Graphics" and to his advisor, Prof. Emmanuel Agu.
Mining Oncology Data: Knowledge Discovery in Clinical Performance of Cancer Patients
John Hayward, Prof. Carolina Ruiz, and Dr. Giles Whales (UMass Medical School)
Poster (pdf)
- This MS thesis focuses on the analysis of clinical oncology data through machine learning and data mining techniques. In our research, we use these techniques to develop models for predicting cancer patient medical outcomes. Oncology data is collected using a clinical database we have developed in conjunction with UMass Medical School. Data analysis in oncology is often constrained by type variation, inconsistent feature availability, high data dimensionality, and small study populations. We compensate for these difficulties using feature selection, meta-learning, and other algorithmic methods during data preparation. Machine learning processes chosen for accuracy, transparency, and medical readability are used to generate our classification models. Classification goals include longevity, quality-of-life, diagnostics, surgical prospects, and patient demographics. We compare the accuracy of classification models for aspects of pancreatic cancers (tumor sizing, nodal involvement, cytology, etc.) against current medical screening techniques to show how data mining may improve predictive oncology diagnosis.
John Hayward grew up in Connecticut, and got his bachelor's in computer science at UMass Amherst. This work at UMass Medical School over the past year has been in fulfillment of his MS degree, and his results are currently lined up for possible publication in both medical and computer science literature. After graduation, he will be working on air traffic control systems at Raytheon's Northeast Software Engineering Center in Marlborough, MA, and he may also pursue further research work with UMass Med part-time.
Wavelet-Based Multiresolution for Mobile Graphics
Fan Wu and Prof. Emmanuel Agu.
Poster (pdf)
- Wavelets and multiresolution analysis provide useful and efficient tools for representing functions at multiple Levels of Detail (LOD). Wavelet representations have been used in a broad range of applications, including computer graphics, image compression and physical simulation. Our research mainly focuses on how to apply wavelets to graphics in mobile environments. Specifically, we focus on transmitting large images and meshes to mobile devices adaptively. Mobile devices such as cell phones, PDAs and laptops have become increasingly popular. Due to their limited resources, such as storage, CPU and battery power, large meshes or images cannot be rendered at full resolution on mobile devices. Wavelet-based multiresolution is a good solution that permits piece-wise transmission and rendering of large graphics models. A system design of this approach is proposed. The implementations and the experimental results of wavelets on 2D images and 3D meshes will be presented. The challenges for future research are discussed as well. The results of the research work may be widely applied in commercial areas, including mobile device manufacturer, mobile gaming and electronic shopping.
Fan Wu came to WPI from China. He received his MS degree from the Department of Computer Science and Technology at Nanjing University of Posts and Telecommunications. The research work presented in the poster is for his PhD degree.
Maintained by webmaster@cs.wpi.eduLast modified: August 03, 2006 08:58:16