In this paper, we introduce a novel approach to time-series prediction realized both at the linguistic and numerical level. It exploits fuzzy cognitive maps (FCMs) along with a recently proposed learning method that takes advantage of real-coded genetic algorithms. FCMs are used for modeling and qualitative analysis of dynamic systems. Within the framework of FCMs, the systems are described by means of concepts and their mutual relationships. The proposed prediction method combines FCMs with granular, fuzzy-set-based model of inputs. One of their main advantages is an ability to carry out modeling and prediction at both numerical and linguistic levels. A comprehensive set of experiments has been carried out with two major goals in mind. One is to assess quality of the proposed architecture, the other to examine the influence of its parameters of the prediction technique on the quality of prediction. The obtained results, which are compared with other prediction techniques using fuzzy sets, demonstrate that the proposed architecture offers substantial accuracy expressed at both linguistic and numerical levels.
In this study, we introduce an advanced architecture of genetically optimized Hybrid Fuzzy Neural Networks (gHFNN) and develop a comprehensive design methodology supporting their construction. A series of numeric experiments is included to illustrate the performance of the networks. The construction of gHFNN exploits fundamental technologies of Computational Intelligence (CI), namely fuzzy sets, neural networks, and genetic algorithms (GAs). The architecture of the gHFNNs results from a synergistic usage of the genetic optimization-driven hybrid system generated by combining Fuzzy Neural Networks (FNN) with Polynomial Neural Networks (PNN). In this tandem, a FNN supports the formation of the condition part of the rule-based structure of the gHFNN. The conclusion part of the gHFNN is designed using PNNs. We distinguish between two types of the simplified fuzzy inference rule-based FNN structures showing how this taxonomy depends upon the type of a fuzzy partition of input variables. As to the conclusion part of the gHFNN, the development of the PNN dwells on two general optimization mechanisms: the structural optimization is realized via GAs whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the gHFNN, we experimented with three representative numerical examples. A comparative analysis demonstrates that the proposed gHFNN come with higher accuracy as well as superb predictive capabilities when compared with other neurofuzzy models.
The article analyzes consecutive phases of time series modelling with Fuzzy Cognitive Maps. The subject of interest are features determining models of good quality. First, we present the procedure: design phase, learning phase, and in the end - application. The discussion is illustrated with experiments on two synthetic time series. We have shown that the design phase determines qualitative and quantitative effectiveness of modelling. We have addressed effects of misdesigns: too large, too small or unfit at all maps on modelling quality.
New methodologies and better techniques are the rule in software engineering, and users of large and complex methodologies benefit greatly from specialized software support tools. However, developing such tools is both difficult and expensive, because developers must implement a lot of functionality in a short time. A promising solution is component-based software development, in particular package-oriented programming (POP). POP fails, however, to satisfy all the requirements of large, complex software engineering tasks. A more generic POP architecture would better serve the development of software engineering environments for large and complex methodologies. Such an architecture emerged from our development experiences with two software engineering research tools: Holmes, a domain analysis support tool; and Egidio, a unified-modeling-language-based business modeling tool. We found this particular architecture simple to understand, easy to implement, and a natural candidate for a generic POP architecture. Our generic architecture satisfies the additional requirements we deem important for larger, more complex software engineering activities. Our experiences show that the strength of this architecture lies in its simplicity and ability to work with multiple users and quickly integrate a wide variety of applications. It is not perfect, but we present it as a first step toward a more general package-oriented architecture to encourage further research in this area.
The rapid expansion of the satellite industry has presented numerous opportunities across various sectors and significantly transformed people's daily lives. However, the high energy consumption resulting from frequent task execution poses challenges for satellite management. Energy consumption has become an important factor to be considered in the design of future satellite management systems. The energy-efficient satellite range scheduling problem (EESRSP) aims to optimize task sequencing profits within the satellite management system while simultaneously conserving energy. To address this problem, a mixed-integer scheduling model is constructed, taking into account the energy consumption of ground stations during telemetry, tracking and command (TT&C) operations. Then, we propose a reinforcement learning-based memetic algorithm (RL-MA) that incorporates a heuristic initialization method (HIM). The HIM enables the algorithm to rapidly generate high-quality initial solutions by leveraging task features associated with EESRSRP. RL-MA employs both population search and local search techniques to explore the satellite TT&C task plan. RL-MA incorporates two genetic operators, crossover and mutation, into the population-based search. In the local search stage, multiple random and heuristic local search operators are incorporated through an ensemble local search strategy (ELSS). To improve search performance, Q-learning, a classical class of reinforcement learning (RL) methods tailored to problem characteristics, is utilized for selecting effective operators. RL dynamically adjusts local search operators based on strategy performance. Experimental results demonstrate that the proposed RL-MA can effectively generate sound solutions for EESRSP with varying task scales. Furthermore, the improvement strategies employed in the algorithm are validated to enhance the scheduling performance of RL-MA. This study reveals that integrating RL with an ensemble of local search operators can significantly enhance the algorithm's exploit capability. Moreover, this local search approach applies to solving other types of satellite scheduling problems.
The task of anomaly detection in data is one of the main challenges in data science because of the wide plethora of applications and despite a spectrum of available methods. Unfortunately, many of anomaly detection schemes are still imperfect i.e., they are not effective enough or act in a non-intuitive way or they are focused on a specific type of data. In this study, the classical method of Isolation Forest is thoroughly analyzed and augmented by bringing an innovative approach. This is k-Means-Based Isolation Forest that allows to build a search tree based on many branches in contrast to the only two considered in the original method. k-Means clustering is used to predict the number of divisions on each decision tree node. As supported through experimental studies, the proposed method works effectively for data coming from various application areas including intermodal transport and geographical, spatio-temporal data. In addition, it enables a user to intuitively determine the anomaly score for an individual record of the analyzed dataset. The advantage of the proposed method is that it is able to fit the data at the step of decision tree building. Moreover, it returns more intuitively appealing anomaly score values.
In this paper, we present results of uncertain state estimation of systems that are monitored with limited accuracy. For these systems, the representation of state uncertainty as confidence intervals offers significant advantages over the more traditional approaches with probabilistic representation of noise. While the filtered-white-Gaussian noise model can be defined on grounds of mathematical convenience, its use is necessarily coupled with a hope that an estimator with good properties in idealised noise will still perform well in real noise. In this study we propose a more realistic approach of matching the noise representation to the extent of prior knowledge. Both interval and ellipsoidal representation of noise illustrate the principle of keeping the noise model simple while allowing for iterative refinement of the noise as we proceed. We evaluate one nonlinear and three linear state estimation technique both in terms of computational efficiency and the cardinality of the state uncertainty sets. The techniques are illustrated on a synthetic and a real-life system.
