The proceedings contain 71 papers. The topics discussed include: soft computing algorithms applied to the segmentation of nerve cell images;patternrecognition based on time-frequency distributions of radar micro-Dopp...
详细信息
ISBN:
(纸本)0769522947
The proceedings contain 71 papers. The topics discussed include: soft computing algorithms applied to the segmentation of nerve cell images;patternrecognition based on time-frequency distributions of radar micro-Doppler dynamics;a quantitative software quality evaluation model for the artifacts of component based development;a new approach to software requirements elicitation;using datamining technology to design an intelligent CIM system for IC manufacturing;datamining for imprecise temporal associations;analysis of breast cancer using datamining and statistical techniques;analyzing the conditions of coupling existence based on program slicing and some abstract information-flow;a study of model layers and reflection;a general scalable implementation of fast matrix multiplication algorithms on distributed memory computers;error prediction for multi-classification;an integer support vector machine;and layered neural networks computations.
This paper presents a shallow dialogue analysis model, aimed at human-human dialogues in the context of staff or business meetings. Four components of the model are defined, and several machinelearning techniques are...
详细信息
ISBN:
(纸本)354024509X
This paper presents a shallow dialogue analysis model, aimed at human-human dialogues in the context of staff or business meetings. Four components of the model are defined, and several machinelearning techniques are used to extract features from dialogue transcripts: maximum entropy classifiers for dialogue acts, latent semantic analysis for topic segmentation, or decision tree classifiers for discourse markers. A rule-based approach is proposed for solving cross-modal references to meeting documents. The methods are trained and evaluated thanks to a common data set and annotation format. The integration of the components into an automated shallow dialogue parser opens the way to multimodal meeting processing and retrieval applications.
Nominal feature is one type of symbolic features, whose feature values are completely unordered. The most often used existing similarity metrics for symbolic features is the Hamming metric, where similarity computatio...
详细信息
A machinelearning technique called Graph-Based Induction (GBI) efficiently extracts typical patterns from graph-structured data by stepwise pair expansion (pairwise chunking). It is very efficient because of its gree...
详细信息
A machinelearning technique called Graph-Based Induction (GBI) efficiently extracts typical patterns from graph-structured data by stepwise pair expansion (pairwise chunking). It is very efficient because of its greedy search. Meanwhile, a decision tree is an effective means of data classification from which rules that are easy to understand can be obtained. However, a decision tree could not be constructed for the data which is not explicitly expressed with attribute-value pairs. This paper proposes a method called Decision Tree Graph-Based Induction (DT-GBI), which constructs a classifier (decision tree) for graph-structured data while simultaneously constructing attributes for classification using GBI. Substructures (patterns) are extracted at each node of a decision tree by stepwise pair expansion in GBI to be used as attributes for testing. Since attributes (features) are constructed while a classifier is being constructed, DT-GBI can be conceived as a method for feature construction. The predictive accuracy of a decision tree is affected by which attributes (patterns) are used and how they are constructed. A beam search is employed to extract good enough discriminative patterns within the greedy search framework. Pessimistic pruning is incorporated to avoid overfitting to the training data. Experiments using a DNA dataset were conducted to see the effect of the beam width and the number of chunking at each node of a decision tree. The results indicate that DT-GBI that uses very little prior domain knowledge can construct a decision tree that is comparable to other classifiers constructed using the domain knowledge. DT-GBI was also applied to analyze a real-world hepatitis dataset as a part of evidence-based medicine. Four classification tasks of the hepatitis data were conducted using only the time-series data of blood inspection and urinalysis. The preliminary results of experiments, both constructed decision trees and their predictive accuracies as well as
The proceedings contain 134 papers. The special focus in this conference is on Ubiquitous intelligence and smart worlds. The topics include: Human activity recognition based on surrounding things;a service management ...
ISBN:
(纸本)3540308032
The proceedings contain 134 papers. The special focus in this conference is on Ubiquitous intelligence and smart worlds. The topics include: Human activity recognition based on surrounding things;a service management system for coordinating smart things in smart spaces;a world model for smart spaces;dealing with emotional factors in agent based ubiquitous group decision;a multi-agent software platform accommodating location-awareness for smart space;application-driven customization of an embedded java virtual machine;intelligent object extraction algorithm based on foreground/background classification;ubiquitous organizational information service framework for large scale intelligent environments;temporal-spatial methodology for application checking of the systems in the ubiquitous environment;multivariate stream data reduction in sensor network applications;implementing a graph neuron array for patternrecognition within unstructured wireless sensor networks;building graphical model based system in sensor networks;the design and implementation of a location-aware service bundle manager in smart space environments;human position/height detection using analog type pyroelectric sensors;semantic based sensor grid and norms enforcement as a coordination strategy in ubiquitous environments.
This paper investigates the automatic analysis and segmentation of meetings. A meeting is analysed in terms of individual behaviours and group interactions, in order to decompose each meeting in a sequence of relevant...
详细信息
ISBN:
(纸本)354024509X
This paper investigates the automatic analysis and segmentation of meetings. A meeting is analysed in terms of individual behaviours and group interactions, in order to decompose each meeting in a sequence of relevant phases, named meeting actions. Three feature families are extracted from multimodal recordings: prosody from individual lapel microphone signals, speaker activity from microphone array data and lexical features from textual transcripts. A statistical approach is then used to relate low-level features with a set of abstract categories. In order to provide a flexible and powerful framework, we have employed a dynamic Bayesian network based model, characterized by multiple stream processing and flexible state duration modelling. Experimental results demonstrate the strength of this system, providing a meeting action error rate of 9%.
Technological advances in experimental and computational molecular biology have revolutionized the whole fields of biology and medicine. Large-scale sequencing, expression and localization data have provided us with a...
详细信息
ISBN:
(纸本)3540305068
Technological advances in experimental and computational molecular biology have revolutionized the whole fields of biology and medicine. Large-scale sequencing, expression and localization data have provided us with a great opportunity to study biology at the system level. I will introduce some outstanding problems in genome expression and regulation network in which better modern statistical and machinelearning technologies are desperately needed. Recent revolution in genomics has transformed life science. For the first time in history, mankind has been able to sequence the entire human own genome. Bioinformatics, especially computational molecular biology, has played a vital role in extracting knowledge from vast amount of information generated by the high throughput genomics technologies. Today, I am very happy to deliver this key lecture at the Firstinternational Conference on patternrecognition and machine Intelligence at the world renowned Indian statistical Institute (ISI) where such luminaries as Mahalanobis, Bose, Rao and others had worked before. And it is very timely that genomics has attracted new generation of talented young statisticians, reminding us the fact that statistics was essentially conceived from and continuously nurtured by biological problems. pattern/rule recognition is at the heart of all learning process and hence of all disciplines of sciences, and comparison is the fundamental method: it is the similarities that allow inferring common rules;and it is the differences that allow deriving new rules. Gene expression, normally referring to the cellular processes that lead to protein production, is controlled and regulated at multiple levels. Cells use this elaborate system of "circuits" and "switches" to decide when, where and by how much each gene should be turned on (activated, expressed) or off (repressed, silenced) in response to environmental clues. Genome expression and regulation refer to coordinated expression and regulation of m
We propose a learning and prediction based paradigm for designing smart home environments. The foundation of this paradigm lies in information theory as it manages uncertainties of the inhabitants' contexts (e.g.,...
详细信息
In this study, a hybrid intelligent datamining methodology, genetic algorithm based support vector machine (GASVM) model, is proposed to explore stock market tendency. In this hybrid datamining approach, GA is used ...
详细信息
This paper summarizes some of the current research challenges arising from multi-channel sequence processing. Indeed, multiple real life applications involve simultaneous recording and analysis of multiple information...
详细信息
ISBN:
(纸本)3540290737
This paper summarizes some of the current research challenges arising from multi-channel sequence processing. Indeed, multiple real life applications involve simultaneous recording and analysis of multiple information sources, which may be asynchronous, have different frame rates, exhibit different stationarity properties, and carry complementary (or correlated) information. Some of these problems can already be tackled by one of the many statistical approaches towards sequence modeling. However, several challenging research issues are still open, such as taking into account asynchrony and correlation between several feature streams, or handling the underlying growing complexity. In this framework, we discuss here two novel approaches, which recently started to be investigated with success in the context of large multimodal problems. These include the asynchronous HMM, providing a principled approach towards the processing of multiple feature streams, and the layered HMM approach, providing a good formalism for decomposing large and complex (multi-stream) problems into layered architectures. As briefly reported here, combination of these two approaches yielded successful results on several multi-channel tasks, ranging from audio-visual speech recognition to automatic meeting analysis.
暂无评论