2017
Workshop on Fuzzy Similarity on December 11, 2017
June 4, 2017
0

IEEE CTS ComSoc/SP/CS/EMBS is collaborating with Texas State University,  School of Engineering on Senior Design Activity event  at San Marcus and offer a workshop on Fuzzy Similarity.

For the recent decades, we have seen rapid progress on hardware and intelligent applications. These bundled systems are widely dependent on traditional Artificial Intelligence techniques such as neural nets and deep learning. However, with a chip that can contain a billion of transistors, it is time to build a “massively parallel array processor” on such a chip. The massively parallel array processor can be used to implement the nearest neighbor (most fuzzy similar) classifiers for the recognition and classification problem, and the function computation model. That is, fuzzy similarity measurements can be used to find the nearest neighbor.

The modern CPU’s “functional programming model” is limited by the need to build functions that are denoted by expressions. The fuzzy-similarity-based function computation model is different. Given x, it searches the most fuzzy similar neighbor m of x and takes the function value f(m) of m as the function value f(x) of x.  This computation model is fundamentally a table-look-up approach. Previously, we did not try this model because a computer CPU is limited. Now, a chip can contain a billion of transistors. Also, we have seen the tremendous growth in silicon photonic circuits.  Therefore, it is time to try this function computation model.

Topics in the 2017 International Workshop on Fuzzy Similarity include but are not limited to the following:

  • Feature Extractions and Fuzzy Similarity Measurements.
  • Applications of   Fuzzy Similarity in image recognition, moving object detection, image description, time data classification, function computation model, self-driving car, natural language processing,  health care, genome, virtualization, and big data analysis.
  • Hardware and Platforms: Massively Parallel Array Processor, FPGA, IC Chips with a massive number of transistors, and OpenCL for heterogeneous computing platforms.
  • MEMS (Micro Electro Mechanical System) or Silicon Photonic circuits for massively parallel processors.

We welcome your participation and contribution. If you wish to present, please send a quick message to Dr. CL Chang nicesoft2009@gmail.com  or  f.behmann@ieee.org