A5024038761- Stock Market Forecasting Methods
- Medical Image Segmentation Techniques
- Magnetic confinement fusion research
- Time Series Analysis and Forecasting
- Superconducting Materials and Applications
- Particle accelerators and beam dynamics
- Advanced Vision and Imaging
- Complex Systems and Time Series Analysis
- Forecasting Techniques and Applications
- Financial Risk and Volatility Modeling
- Computer Graphics and Visualization Techniques
Cornell University
2023
University of Washington
2002
The KSTAR (Korea Superconducting Tokamak Advanced Research) project is the major effort of Korean National Fusion Program to design, construct, and operate a steady-state-capable superconducting tokamak. led by Korea Basic Science Institute shared national laboratories, universities, industry along with international collaboration. It in conceptual design phase aims for first plasma mid 2002. key features are: radius 1.8 m, minor 0.5 toroidal field 3.5 T, current 2 MA, flexible shaping...
In this study, we introduce SimStock, a novel framework leveraging self-supervised learning and temporal domain generalization techniques to represent similarities of stock data. Our model is designed address two critical challenges: 1) distribution shift (caused by the non-stationarity financial markets), 2) ambiguity in conventional regional sector classifications (due rapid globalization digitalization). SimStock exhibits outstanding performance identifying similar stocks across four...
A description is given of the computer graphics aspects two architectures designed for imaging and graphics. The systems use parallel pipelined high-performance operations. UWGPSP3 uses only commercially available off-the-shelf chips, consists a TM34020 system processor four TMS34082 floating point coprocessors that can be configured into or SIMD modes depending on algorithm. UWGSP4 dedicated ASIC chips higher performance, main computational parts: vector with 16 processing units, used...