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

This chapter discusses the application of the innovative methods described in Chap. 5 and their experimental evaluation. The methods were tested to evaluate the different biometric aspects of the implemented techniques, including accuracy, robustness, speed, cost, interoperability, usability, acceptability, security, and privacy.

Application and Experimental Evaluation of Methods

In this paper we present a study on the feasibility of the prediction of the solar radiation on a location giving the meteorological measurement in surrounding locations on a mesoscale system scale. The data from four public stations run by the Lombardy regional agency for environmental protection (ARPA) have been used as dataset for training a neural network in order to predict with one-hour lag the global radiation in one station using the data from the other three stations. The results have been compared with other two models: the first makes use of only the data from the station to be predicted, while the second exploits all the available information considering all the four stations as input sources. The dataset has been formed using data from the ARPA stations in Milano, Crema, Osio sotto, e Cassano d'Adda, considering the years 2002–2007.

Multi-point solar prediction through feed-forward neural networks

This paper presents an automatic design flow for digital special purpose feed-forward multi-layer neural networks (SPNNs). The target architecture is constituted by a collection of basic elements called Pseudo-Neurons (PNs) connected in a pipelined-tree manner, to suit efficient VLSI implementation and, due to an internal pipelined multiprocessing, achieving low latency and good throughput. The related CAD environment implements three successive stages: the weights discretization (WD), the architectural synthesis (AS), and the VHDL model generation (VHDL_G).

From Behavior to VHDL: A CAD Environment for SPNNs

Biliary Atresia (BA) refers to a disease that mostly affects newborns by partially obstructing the bile ducts from the liver to the intestines, causing the trapped bile to damage the liver itself and often resulting in the need for a transplant. To detect BA, expert personell (e.g., pediatricians) or non-experts (e.g., the parents) usually analyze the color of the feces with the help of a reference stool color card. To automate this process, some approaches in the literature proposed smartphone apps enabling the parents to capture an image of the feces, select a point of the image to analyze, and compare it with the stool color card. However, such approaches consider only the local pixel chosen for matching and are therefore highly dependent on the position chosen by the user, who may choose a non-significant pixel to perform the analysis. In this work, we propose the first method in the literature for BA detection that considers a color-based segmentation and a nearest neighbor classification. Differently than the approaches in the literature, the color segmentation clusters the image in different areas based on the color and permits to automatically and robustly consider the corresponding cluster, and not only the local pixel, to perform the classification. Results on a database captured in uncontrolled conditions show the validity of the approach.

Biliary Atresia Detection Using Color Clustering and Nearest Neighbor Classification: A User Interactive Approach

Special Issue on Artificial Intelligence-of-Things (AIoT): Opportunities, Challenges, and Solutions-Part II: Artificial-Intelligence-Powered Internet of Things

Vincenzo Piuri

Papers

Application and Experimental Evaluation of Methods

Multi-point solar prediction through feed-forward neural networks

From Behavior to VHDL: A CAD Environment for SPNNs

Biliary Atresia Detection Using Color Clustering and Nearest Neighbor Classification: A User Interactive Approach

Special Issue on Artificial Intelligence-of-Things (AIoT): Opportunities, Challenges, and Solutions-Part II: Artificial-Intelligence-Powered Internet of Things