In this paper, we present a novel active learning strategy, named dynamic active learning with SVM to improve the effectiveness of learning sample selection in active learning. The algorithm is divided into two steps....
详细信息
Since the introduction of rough sets in 1982 by Professor Zdzislaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of ...
详细信息
Since the introduction of rough sets in 1982 by Professor Zdzislaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of forms such as knowledge discovery, approximate reasoning, intelligent and multiagent system design, knowledge intensive computations .
It seems characteristic for humans to detect structural patterns in the world to anticipate future states. Therefore, scientific and common sense cognition could be described as information processing which infers rul...
详细信息
ISBN:
(数字)9783540742623
ISBN:
(纸本)9783540742616
It seems characteristic for humans to detect structural patterns in the world to anticipate future states. Therefore, scientific and common sense cognition could be described as information processing which infers rule-like laws from patterns in data-sets. Since information processing is the domain of computers, artificial cognitive systems are generally designed as pattern discoverers. This paper questions the validity of the information processing paradigm as an explanation for human cognition and a design principle for artificial cognitive systems. Firstly, it is known from the literature that people suffer from conditions such as information overload, superstition, and mental disorders. Secondly, cognitive limitations such as a small short-term memory, the set-effect, the illusion of explanatory depth, etc. raise doubts as to whether human information processing is able to cope with the enormous complexity of an infinitely rich (amorphous) world. It is suggested that, under normal conditions, humans construct information rather than process it. The constructed information contains anticipations which need to be met. This can be hardly called information processing, since patterns from the "outside" are not used to produce action but rather to either justify anticipations or restructure the cognitive apparatus. When it fails, cognition switches to pattern processing, which, given the amorphous nature of the experiential world, is a lost cause if these patterns and infer-red rules do not lead to a (partial) reorganisation of internal structures such that constructed anticipations can be met again. In this scenario, superstition and mental disorders are the result of a profound and/or random restructuring of already existing cognitive components (e.g., action sequences). This means that whenever a genuinely cognitive system is exposed to pattern processing it may start to behave superstitiously. The closer we get to autonomous self-motivated artificial cognitive sy
In this paper we describe a datamining approach for detection of anomalous vessel behaviour. The suggested approach is based on Bayesian networks which have two important advantages compared to opaque machine learnin...
详细信息
In this paper we describe a datamining approach for detection of anomalous vessel behaviour. The suggested approach is based on Bayesian networks which have two important advantages compared to opaque machinelearning techniques such as neural networks: (1) possibility to easily include expert knowledge into the model, and (2) possibility for humans to understand and interpret the learned model. Our approach is implemented and tested on synthetic data, where initial results show that it can be used for detection of single-object anomalies such as speeding.
Solar imaging is currently an active area of research. A fast hybrid system for the automated detection and classification of sunspot groups on MDI Continuum images using active regions data extracted from MDI Magneto...
详细信息
Solar imaging is currently an active area of research. A fast hybrid system for the automated detection and classification of sunspot groups on MDI Continuum images using active regions data extracted from MDI Magnetogram images is presented in this paper. The system has three major stages: sunspots detection from MDI Continuum images, sunspots grouping and Mcintosh classification of sunspot groups. Image processing and machinelearning are integrated in all these stages.
The need for robust health monitoring and prognostics of structural components in remote or difficult-to-access locations, e.g. helicopter rotor-head structure, is driving the advancement of wireless intelligent senso...
详细信息
The need for robust health monitoring and prognostics of structural components in remote or difficult-to-access locations, e.g. helicopter rotor-head structure, is driving the advancement of wireless intelligent sensor devices (WISD). Damage detection techniques, combined with advanced signal processing, are the core components of a structural health monitoring (SHM) system. In this context, feature extraction is an essential component of a SHM system that converts raw sensor data into useful information about the structure health condition. The level of signal processing that can be performed in a WISD depends on the capability of the processing element in terms of speed, memory and energy consumption. But the real bottleneck for energy efficiency is the fact that communications dominate the WISD energy consumption. Therefore, running intelligent local data interrogation algorithms on-board the WISD is a mechanism through which considerable battery power can be preserved. In that sense, in this paper a soft histogram feature extraction algorithm is developed to extract damage-sensitive information from measured response data of tie-bar component of the main rotor hub of a Lynx helicopter. In addition, a method for patternrecognition and critical degradation detection of tie-bar is proposed based on the extracted features and a combination of statistical process control and fuzzy sets theory. Results show the applicability of the proposed approaches.
Since the introduction of rough sets in 1982 by Professor Zdzisaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of...
Since the introduction of rough sets in 1982 by Professor Zdzisaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of forms such as knowledge discovery, approximate reasoning, intelligent and multiagent system design, knowledge intensive computations. The cutting-edge knowledge technologies have great impact on learning, patternrecognition, machine intelligence and automation of acquisition, transformation, communication, exploration and exploitation of knowledge. A principal thrust of such technologies is the utilization of methodologies that facilitate knowledge processing. To present the state-of-the-art scientific results, encourage academic and industrial interaction, and promote collaborative research in rough sets and knowledge technology worldwide, the 3rdinternationalconference on Rough Sets and Knowledge Technology will be held in Chengdu, China, May 17～19, 2008. It will provide a forum for researchers to discuss new results and exchange ideas, following the successful RSKT'06 (Chongqing, China) and JRS'07 (RSKT'07 together with RSFDGrC'07) (Toronto, Canada).
Since the introduction of rough sets in 1982 by Professor Zdzisaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of...
详细信息
Since the introduction of rough sets in 1982 by Professor Zdzisaw Pawlak, we have witnessed great advances in both theory and applications. Rough set theory is closely related to knowledge technology in a variety of forms such as knowledge discovery, approximate reasoning, intelligent and multiagent system design, knowledge intensive computations. The cutting-edge knowledge technologies have great impact on learning, patternrecognition, machine intelligence and automation of acquisition, transformation, communication, exploration and exploitation of knowledge. A principal thrust of such technologies is the utilization of methodologies that facilitate knowledge processing. To present the state-of-the-art scientific results, encourage academic and industrial interaction, and promote collaborative research in rough sets and knowledge technology worldwide, the 3rdinternationalconference on Rough Sets and Knowledge Technology will be held in Chengdu, China, May 17～19, 2008. It will provide a forum for researchers to discuss new results and exchange ideas, following the successful RSKT'06 (Chongqing, China) and JRS'07 (RSKT'07 together with RSFDGrC'07) (Toronto, Canada).
There are several approaches in trying to solve the Quantitative Structure-Activity (QSAR) problem. These approaches are based either on statistical methods or on predictive datamining. Among the statistical methods,...
详细信息
ISBN:
(纸本)9781424401956
There are several approaches in trying to solve the Quantitative Structure-Activity (QSAR) problem. These approaches are based either on statistical methods or on predictive datamining. Among the statistical methods, one should consider regression analysis, patternrecognition (such as cluster analysis, factor analysis and principal components analysis) or partial least squares. Predictive datamining techniques use either neural networks, or genetic programming, or neuro-fuzzy knowledge. These approaches have a low explanatory capability or non at all. This paper attempts to establish a new approach in solving QSAR problems using descriptive datamining. This way, the relationship between the chemical properties and the activity of a substance would be comprehensibly modeled.
The low error rate of Naive Bayes (NB) classifier has been described as surprising. It is known that class conditional independence of the features is sufficient but not a necessary condition for optimality of NB. Thi...
详细信息
ISBN:
(纸本)9781424401956
The low error rate of Naive Bayes (NB) classifier has been described as surprising. It is known that class conditional independence of the features is sufficient but not a necessary condition for optimality of NB. This study is about the difference between the estimated error and the true error of NB taking into account feature dependencies. Analytical results are derived for two binary features. Illustration examples are also provided.
暂无评论