Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

An Introduction to MultiAgent Systems.

Artificial neural networks (ANNs) constitute a class of flexible nonlinear models designed to mimic biological neural systems. In this entry, we introduce ANN using familiar econometric terminology and provide an overview of ANN modeling approach and its implementation methods. † Correspondence: Chung-Ming Kuan, Institute of Economics, Academia Sinica, 128 Academia Road, Sec. 2, Taipei 115, Taiwan; ckuan@econ.sinica.edu.tw. †† I would like to express my sincere gratitude to the editor, Professor Steven Durlauf, for his patience and constructive comments on early drafts of this entry. I also thank Shih-Hsun Hsu and Yu-Lieh Huang for very helpful suggestions. The remaining errors are all mine.

Artificial neural networks

International Technology Roadmap for Semiconductors 2003の要求清浄度について － シリコンウエハ表面と雰囲気環境に要求される清浄度, 分析方法の現状について －

This chapter introduces the subject of statistical pattern recognition (SPR). It starts by considering how features are defined and emphasizes that the nearest neighbor algorithm achieves error rates comparable with those of an ideal Bayes’ classifier. The concepts of an optimal number of features, representativeness of the training data, and the need to avoid overfitting to the training data are stressed. The chapter shows that methods such as the support vector machine and artificial neural networks are subject to these same training limitations, although each has its advantages. For neural networks, the multilayer perceptron architecture and back-propagation algorithm are described. The chapter distinguishes between supervised and unsupervised learning, demonstrating the advantages of the latter and showing how methods such as clustering and principal components analysis fit into the SPR framework. The chapter also defines the receiver operating characteristic, which allows an optimum balance between false positives and false negatives to be achieved.

Statistical Pattern Recognition

System og fremgangsmåde til overvågning af lasersvejsninger og til at give en indikation af svejsningens kvalitet

The FERMI system has been designed to perform acquisition and DSP of calorimeter data in high energy collision experiments, planned at the LHC collider (CERN, Geneva, CH). The system relies mainly upon the FERMI chip, a dedicated VLSI multi-chip device: it requires rad-hard features due to the critical application, to the difficulties in maintenance. And to the operating environment. In this paper, a modular approach to support policies for self diagnosis and fault tolerance is discussed. >

A modular fault-tolerant approach to design a front-end microsystem for calorimetry at LHC

Several DSP algorithms need to remove high-frequency or impulsive noise while preserving edges, e.g., in speech and image processing applications: median filtering has been proved more effective to achieve this goal than other filtering techniques. Efficient architectural implementation for real-time applications involves a careful VLSI design, which takes into account modularity, regularity, adaptability, throughput, circuit complexity, and fault tolerance. Two new architectural approaches are presented in this paper to deal with different application and implementation constraints: the pipelined polarizing median filter and the pipelined sorting median filter. >

https://ieeexplore.ieee.org/iel2/1120/8083/00352171.pdf

Pipelined median filters

Spectral warping (SW) is the digital signal processing (DSP) procedure of transforming an original digital sequence to a new one having special spectral properties: equally spaced samples of its DFT (Discrete Fourier Transform) are identical to the unequally-spaced frequency samples of DFT of the original sequence. The use of SW can open up new opportunities in test signal generation as well as in test response analysis for mixed signal circuits.

Using spectral warping for instrumentation and measurements in mixed-signal testing

Modern distributed file system realizations offer only partially resource location transparency, resource location independence, fault tolerance, load balancing, heterogeneity, self-configuration, and simplified user access. Traditional portability techniques developed in these systems become unsuited in highly dynamic environments.

To solve these problems within a homogeneous framework we studied and experimented the use of static and mobile agents in a portable environment. In this paper we describe the philosophy, the structure, and the prototype realization of the Agent-based Distributed File System (ADFS). The main properties of this innovative distributed file system are resource location transparency, resource location independence, self-configuration, and heterogeneity of the underlying hardware and operating system architectures.

Vincenzo Piuri

Papers

System og fremgangsmåde til overvågning af lasersvejsninger og til at give en indikation af svejsningens kvalitet

A modular fault-tolerant approach to design a front-end microsystem for calorimetry at LHC

Pipelined median filters

Using spectral warping for instrumentation and measurements in mixed-signal testing

An Agent-Based Approach to Full Interoperability and Allocation Transparency in Distributed File Systems