The design of information granules assumes a central position in the discipline of Granular Computing and its applications. The principle of justifiable granularity offers a conceptually and algorithmically attractive way of designing information granule completed on a basis of some experimental evidence (especially present in the form of numeric data). This paper builds upon the existing principle and presents its significant generalization, referred here as an adaptive principle of justifiable information granularity. The method supports a granular data aggregation producing an optimal information granule (with the optimality expressed in terms of the criteria of coverage and specificity commonly used when characterizing quality of information granules). The flexibility of the method stems from an introduction of the adaptive weighting scheme of the data leading to a vector of weights used in the construction of the optimal information granule. A detailed design procedure is provided along with the required optimization vehicle (realized with the aid of the population-based optimization techniques, such as particle swarm optimization and differential evolution). Two direct application areas in which the principle becomes of direct usage include prediction of time series and prediction of spatial data. In both cases, it is advocated that the results formed by the principle are reflective of the precision (quality) of the prediction process.
The study is devoted to the paradigm of rule based computing involving granular information. By information granules we mean a general category of data embracing not only numeric entities (inputs) but any granules (such as intervals or fuzzy sets, in general) being regarded as inputs in the rule-based system. We investigate several categories of models of granularity propagation starting from those based on the use of the mechanisms of possibility and necessity theory, especially possibility and possibility-necessity mechanisms. We also consider the models relying on the use of auxiliary regression models. These models are constructed on the basis of some experimental granular data. A thorough comparative analysis of the introduced models is carried out as well.
Radial basis function (RBF) neural networks form an essential category of architectures of neurocomputing. They exhibit interesting and useful properties of stable and fast learning associated with significant generalization capabilities. This successful performance of RBF neural networks can be attributed to the use of a collection of properly selected RBFs. In this way this category of the networks strongly relies on some domain knowledge about a classification problem at hand. Following this vein, this study introduces fuzzy clustering, and fussy isodata, in particular, as an efficient tool aimed at constructing receptive fields of RBF neural networks. It is shown that the functions describing these fields are completely derived as a by‐product of fuzzy clustering and do not require any further tedious refinements. The efficiency of the design is illustrated with the use of synthetic two‐dimensional data as well as real‐world highly dimensional ECG patterns. The classification of the latter data set clearly points out advantages of RBF neural networks in pattern recognition problems.
Estimation of effort/cost required for development of software products is inherently associated with uncertainty. In this paper, we are concerned with a fuzzy set-based generalization of the COCOMO model (f-COCOMO). The inputs of the standard COCOMO model include an estimation of project size and an evaluation of other parameters. Rather than using a single number, the software size can be regarded as a fuzzy set (fuzzy number) yielding the cost estimate also in form of a fuzzy set. The paper includes detailed results with this regard by relating fuzzy sets of project size with the fuzzy set of effort. The analysis is carried out for several commonly encountered classes of membership functions (such as triangular and parabolic fuzzy sets). The issue of designer-friendliness of the f-COCOMO model is discussed in detail. Here we emphasize a way of propagation of uncertainty and ensuing visualization of the resulting effort (cost). Furthermore we augment the model by admitting software systems to belong partially to the three main categories (namely embedded, semidetached and organic) and discuss key implications of this generalization and highlight its links with a generalized sensitivity analysis. The experimental part of the study illustrates the approach and contrasts it with the standard numeric version of the COCOMO model.
The agile Earth observation satellite scheduling problem (AEOSSP) with time-dependent transition time is a combinatorial optimization challenge. Due to its NP-hardness, problem-tailored methods are sensitive to instances and require massive computational overhead. Recently, deep reinforcement learning (DRL) models have shown promise in efficiently addressing the AEOSSP. However, these models may make decision mistakes in specific scenarios due to prioritizing maximizing average reward expectation over individual decision accuracy during DRL training, directly leading to resource wastage. To address these issues, we propose a reconstruction model (RCM), which is a DRL-based two-stage construction model (CM), including a CM and a reconstruction mechanism (RM). RCM constructs solutions initially using a DRL-trained CM, which are subsequently refined by RM. CM utilizes a more efficient network for policy representation to make decisions. RM applies two operators, "repair" and "removal," with a "repair-removal-repair" solution reconstruction process to identify and rectify decision mistakes from CM, offering a modular component to enhance the stability and solution quality. Experimental results demonstrate that the proposed RCM outperforms the state-of-the-art AEOSSP iterative search method, achieving such performance within a computational time of 0.1 s. Additionally, CM surpasses the state-of-the-art DRL policy model and RM can effectively rectify decision errors or suboptimalities, underscoring its effectiveness in enhancing DRL outcomes.
In many practical situations, we are faced with a necessity to combine sophisticated mathematical knowledge about the analyzed systems with informal expert knowledge. To make this combination natural, it is desirable to reformulate the abstract mathematical knowledge in understandable intuitive terms. In this paper, we show how this can be done for an abstract metric.\nOne way to define a metric is to pick certain properties P1, ..., Pn, and to define a similarity between two objects x and y as the degree to which P1(x) is similar to P1(y) and P2(x) is similar to P2(y), etc.\nSimilarity is naturally described by 1-|d1-d2| (we can use robustness arguments to get this expression). Since we can have infinitely many properties, we should use min for "and". The distance is then 1-this similarity. The resulting metrics are "natural".\nIt seems, at first glance, that not all metrics are natural in this sense. Interestingly, an arbitrary continuous metric can be thus described.\nSimilarly, we can thus describe all "kinematic metrics" (space-time analogues of metrics), while probabilistic explanation is difficult.
Article Share on Application of computational intelligence techniques in active networks Authors: Athanasios V. Vasilakos Institute of Computer Science, Foundation of Research and Technology-Hellas Institute of Computer Science, Foundation of Research and Technology-HellasView Profile , Kostas G. Anagnostakis Distributed Systems Lab, CIS Department, University of Pennsylvania Distributed Systems Lab, CIS Department, University of PennsylvaniaView Profile , Witold Pedrycz Department of Electrical and Computer Engineering, University of Alberta, Canada Department of Electrical and Computer Engineering, University of Alberta, CanadaView Profile Authors Info & Claims SAC '01: Proceedings of the 2001 ACM symposium on Applied computingMarch 2001Pages 448–455https://doi.org/10.1145/372202.372404Published:01 March 2001Publication History 1citation547DownloadsMetricsTotal Citations1Total Downloads547Last 12 Months2Last 6 weeks1 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Publisher SiteGet Access
This paper is concerned with a development of a segmentation technique for electrocardiogram (ECG) signals. Such segmentation is aimed at a lossy signal compression in which each segment can be captured by a simple geometric construct such as, e.g., a linear or quadratic function. The crux of the proposed construct lies in the determination of the optimal segments of data over which they exhibit the highest possible monotonicity (or lowest variability) of the ECG signal. In this sense, the proposed approach generalizes a fundamental and commonly encountered problem of function (data) linearization. The segments are genetically developed using a standard technique of genetic algorithms (GAs). The two fundamental GA constructs, namely a topology of a chromosome and a fitness function governing the optimization process are discussed in detail. The chromosome being coded as a series of floating point numbers contains the endpoints of the segments (segmentation points). The fitness function to be maximized quantifies a level of monotonicity of the ECG data encountered within the segments and takes into consideration differences between the extreme values (minimum and maximum) of its derivatives. As a result of the genetic optimization, we build segments of ECG signals encompassing monotonic (increasing or decreasing) regions of the signal exhibiting a minimal level of variability. A series of experiments dealing with several classes of ECG signals (namely, normal, left bundle branch block beat, and right bundle branch block beat) visualize the effectiveness of the approach and shows the specificity of the linear segments of data. Furthermore, we elaborate on the relationship between the values of the fitness function and the approximation capabilities (quantified by a sum of squared errors between the local model and the data) of the segments of the signal and show that these two descriptors are highly related.
