this paper presents a novel unsupervised fuzzy clustering method based on clonal selection algorithm for anomaly intrusion detection in order to solve the problem of fuzzy k-means algorithm which is particularly sensi...
详细信息
ISBN:
(纸本)3540335846
this paper presents a novel unsupervised fuzzy clustering method based on clonal selection algorithm for anomaly intrusion detection in order to solve the problem of fuzzy k-means algorithm which is particularly sensitive to initialization and fall easily into local optimization. this method can quickly obtain the global optimal clustering with a clonal operator which combines evolutionary search, global search, stochastic search and local search, then detect abnormal network behavioral patterns with a fuzzy detection algorithm. Simulation results on the data set KDD CUP99 show that this method can efficiently detect unknown intrusions with lower false positive rate and higher detection rate.
We investigate the use of independent component analysis (ICA) for speech feature extraction in digits speech recognition systems. We observe that this may be true for recognition tasks based on Geometrical learning w...
详细信息
ISBN:
(纸本)3540335846
We investigate the use of independent component analysis (ICA) for speech feature extraction in digits speech recognition systems. We observe that this may be true for recognition tasks based on Geometrical learning with little training data. In contrast to image processing, phase information is not essential for digits speech recognition. We therefore propose a new scheme that shows how the phase sensitivity can be removed by using an analytical description of the ICA-adapted basis functions. Furthermore, since the basis functions are not shift invariant, we extend the method to include a frequency-based ICA stage that removes redundant time shift information. the digits speech recognition results show promising accuracy. Experiments show that the method based on ICA and Geometrical learning outperforms HMM in a different number of training samples.
this paper presents a novel approach for adaptive online multi-stroke sketch recognition based on Hidden Markov Model (HMM). the method views the drawing sketch as the result of a stochastic process that is governed b...
详细信息
ISBN:
(纸本)3540335846
this paper presents a novel approach for adaptive online multi-stroke sketch recognition based on Hidden Markov Model (HMM). the method views the drawing sketch as the result of a stochastic process that is governed by a hidden stochastic model and identified according to its probability of generating the output. To capture a user's drawing habits, a composite feature combining both geometric and dynamic characteristics of sketching is defined for sketch representation. To implement the stochastic process of online multi-stroke sketch recognition, multi-stroke sketching is modeled as an HMM chain while the strokes are mapped as different HMM states. To fit the requirement of adaptive online sketch recognition, a variable state-number determining method for HMM is also proposed. the experiments prove boththe effectiveness and efficiency of the proposed method.
In this paper, an effective method to discover repeating pattern from audio is proposed. Since the previous feature extraction methods are usually process monophony audio, for extracting more descriptive features from...
详细信息
ISBN:
(纸本)3540335846
In this paper, an effective method to discover repeating pattern from audio is proposed. Since the previous feature extraction methods are usually process monophony audio, for extracting more descriptive features from polyphony audio, Gabor filters bank is introduced. Meanwhile the measure criteria is suggested for qualitatively and quantitatively weighting the discernibility of extracted features. In addition, the presented algorithm is based on the incremental match and has time complexity O(nlog(n)). Experimental evaluations show that our proposed method could extract complete and meaningful repeating patterns from polyphony audio.
To develop a good evaluation of the classes present on a scene is one of the difficulties in patternrecognition. To suitably describe those classes it is necessary to find feature spaces,which allow distinguishing be...
详细信息
ISBN:
(纸本)0780395840
To develop a good evaluation of the classes present on a scene is one of the difficulties in patternrecognition. To suitably describe those classes it is necessary to find feature spaces,which allow distinguishing between them. In this work, we propose an unsupervised segmentation/classification technique based on wavelet textural analysis and self-organizing maps clustering.
Rule induction is a datamining process for acquiring knowledge in terms of symbolic decision rules that explain the data in terms of causal relationship between conditional factors and a given decision/outcome. We pr...
详细信息
ISBN:
(纸本)3540335846
Rule induction is a datamining process for acquiring knowledge in terms of symbolic decision rules that explain the data in terms of causal relationship between conditional factors and a given decision/outcome. We present a Decision Rule Acquisition Workbench (DRAW) that discovers symbolic decision rules, in CNF form, from un-annotated data-sets. Our rule-induction strategy involves three phases: (a) conceptual clustering to cluster and generate a conceptual hierarchy of the data-set;(b) rough sets based rule induction algorithm to generate decision rules from the emergent data clusters;and (c) attribute oriented induction to generalize the derived decision rules to yield high-level decision rules and a minimal rule-set size. We evaluate DRAW with five standard machinelearningdatasets and apply to derive decision rules to understand optic nerve images in the realm of glaucoma decision support.
the proceedings contain 68 papers. the topics discussed include: a comparative analysis of data distribution methods in an agent-based neural system for classification tasks;stochastic differential portfolio games wit...
详细信息
ISBN:
(纸本)0769526624
the proceedings contain 68 papers. the topics discussed include: a comparative analysis of data distribution methods in an agent-based neural system for classification tasks;stochastic differential portfolio games with regime switching model;extracting symbolic rules from clustering of gene expression data;a novel microarray gene selection method based on consistency;combining greedy method and genetic algorithm to identify transcription factor binding sites;investigation of a new artificial immune system model applied to patternrecognition;RLM: a new method of encoding weights in DNA strands;shape representation and distance measure based on retational graph;fast modeling of curved object from two images;research on an improved gray gradient orientation algorithm in anisotropic high-pass filtering;and image color reduction based on self-organizing maps and growing self-organizing neural networks.
the aims and objectives of datamining is to discover actionable knowledge of main interest to real user needs, which is one of Grand Challenges in KDD. Most extant datamining is a data-driven trial-an-error process....
详细信息
ISBN:
(纸本)9781586036157
the aims and objectives of datamining is to discover actionable knowledge of main interest to real user needs, which is one of Grand Challenges in KDD. Most extant datamining is a data-driven trial-an-error process. patterns discovered via predefined models in the above process are often of limited interest to constraint-based real business. In order to work out patterns really interesting and actionable to the real world, pattern discovery is more likely to be a domain-driven human-machine-cooperated process. this talk proposes a practical datamining methodology named "domain-driven datamining". the main ideas include a Domain-Driven In-Depthpattern Discovery framework (DDID-PD), constraint-based mining, in-depthmining, human-cooperated mining and loop-closed mining. Guided by this methodology, we demonstrate some of our work in identifying useful correlations in real stock markets, for instance, discovering optimal trading rules from the existing rule classes, and mining trading rule-stock correlations in stock exchange data. the results have attracted strong interest from both traders and researchers in stock markets. It has shown that the methodology is potential for guiding deep mining of patterns interesting to real business.
this paper explores the customized learning from specific to general for classification learning. Our novel learning framework called SUPE customizes its learning process to the instance to be classified called query ...
详细信息
ISBN:
(纸本)3540335846
this paper explores the customized learning from specific to general for classification learning. Our novel learning framework called SUPE customizes its learning process to the instance to be classified called query instance. the data representation in SUPE is also customized to the query instance. Given a query instance, the training data is transformed into a query matrix, from which useful patterns are discovered for learning. the final prediction of the class label is performed by combining some statistics of the discovered useful patterns. We show that SUPE conducts the search from specific to general in a significantly reduced hypothesis space. the query matrix also facilitates the complicated operations in SUPE. the experimental results on benchmark data sets are encouraging.
Clustering is one branch of unsupervised machinelearningtheory, which has a wide variety of applications in patternrecognition, image processing, economics, document categorization, web mining, etc. Today, we const...
详细信息
ISBN:
(纸本)3540465359
Clustering is one branch of unsupervised machinelearningtheory, which has a wide variety of applications in patternrecognition, image processing, economics, document categorization, web mining, etc. Today, we constantly face how to handle a large number of similar data items, which drives many researchers to contribute themselves to this field. Support vector machine provides a new pathway for clustering, however, it behaves bad in handling massive data. As an emergent theory, artificial immune system can effectively recognize antigens and produce the memory antibodies. this mechanism is constantly used to achieve representative or feature data from raw data. A combinational clustering method is proposed in this paper based on artificial immune system and support vector machine, Experimentation in functionality and performance is done in detail. Finally a more challenging application in elevator industry is conducted. the results strongly indicate that this combinational clustering in this paper is of feasibility and of practice.
暂无评论