Application of a niched Pareto genetic algorithm for selecting features for nuclear transients classification

Abstract

Feature selection for transient classification is the problem of choosing among several monitored parameters (i.e., the features) to be used for efficiently recognizing the developing transient patterns. It is a critical issue for the application of "on condition" diagnostic techniques in complex systems, such as the nuclear power plants, where hundreds of parameters are measured. Indeed, irrelevant and noisy features have been shown to unnecessarily increase the complexity of the classification problem and degrade the diagnostic performance. In this paper, the problem of selecting the features to be used for efficient transient classification is tackled by means of multiobjective genetic algorithms. The approach leads to the identification of a family of equivalently optimal subsets of features, in the Pareto sense. However, difficulties in the convergence of the standard Pareto-based multiobjective genetic algorithm search in large feature spaces may arise in terms of representativeness of the identified Pareto front whose elements may turn out to be unevenly distributed in the objective functions space, thus not providing a full picture of the potential Pareto-optimal solutions. To overcome this problem, a niched Pareto genetic algorithm is embraced in this work. The performance of the feature subsets examined during the search is evaluated in terms of two optimization objectives: the classification accuracy of a Fuzzy K-Nearest Neighbors classifier and the number of features in the subsets. During the genetic search, the algorithm applies a controlled "niching pressure" to spread out the population in the search space so that convergence is shared on different niches of the Pareto front, which is thus evenly covered. The method is tested on a diagnostic problem characterized by a very large number of process features available for the classification of simulated transients in the feedwater system of a boiling water reactor. The dynamics of the transient signals is captured by wavelet decomposition, which actually increases the complexity of the search for the optimal feature subsets by triplicating the number of features to be considered. C 2008 Wiley Periodicals, Inc.

Hydrological cycling optimization‐based multiobjective feature‐selection method for customer segmentation

“…Their proposed approach demonstrated a low error rate and solution count. Baraldi et al 31 proposed a multiobjective GA‐based wrapper method for nuclear‐transient classification. They employed a search strategy designed to release the convergence pressures by different niches of the Pareto front.…”

Performance evaluation of evolutionary multiobjective optimization algorithms for multiobjective fuzzy genetics-based machine learning

“…2. For example, the computation load in each MoGFS study for classification problems was 50 9 1,000 in Ishibuchi et al (2001), 50 9 10,000 in Ishibuchi and Yamamoto (2004), 150 9 1,000 in Setzkorn and Paton (2005), 200 9 5,000 in Ishibuchi and Nojima (2007a), 1,000 9 1,000 in Pulkkinen and Koivisto (2008), and 200 9 1,000 in Baraldi et al (2009). These discussions suggest that the reported classification performance of MoGFS in the literature can be improved by increasing the computation load of EMO algorithms and/or improving their search ability.…”

Homogeneous Continuous-Time, Finite-State Hidden Semi-Markov Modeling for Enhancing Empirical Classification System Diagnostics of Industrial Components

“…To do this, we rely on the strategy discussed by Cannarile et al [25]. (2) The selection of an optimal subset of relevant features to be used for the classification [26] through the scheme proposed in our earlier work [22], i.e., the feature selector behaves as a wrapper around the specific learning algorithm used to construct the classifier [16]. The objective functions used for evaluating and comparing the feature subsets during the search are the recognition rate achieved by the ECS (to be maximized) and the number of features forming the subsets (to be minimized).…”