Byung-In Choi

Interval type-2 fuzzy membership function generation methods for pattern recognition

Type-2 fuzzy sets (T2 FSs) have been shown to manage uncertainty more effectively than T1 fuzzy sets (T1 FSs) in several areas of engineering [4,6-12,15-18,21-27,30]. However, computing with T2 FSs can require undesirably large amount of computations since it involves numerous embedded T2 FSs. To reduce the complexity, interval type-2 fuzzy sets (IT2 FSs) can be used, since the secondary memberships are all equal to one [21]. In this paper, three novel interval type-2 fuzzy membership function (IT2 FMF) generation methods are proposed. The methods are based on heuristics, histograms, and interval type-2 fuzzy C-means. The performance of the methods is evaluated by applying them to back-propagation neural networks (BPNNs). Experimental results for several data sets are given to show the effectiveness of the proposed membership assignments.

Interval Type-2 Fuzzy Membership Function Design and its Application to Radial Basis Function Neural Networks

Type-2 fuzzy sets has been shown to manage uncertainty more effectively than type-1 fuzzy sets in several pattern recognition applications. However, computing with type-2 fuzzy sets can require high computational complexity since it involves numerous embedded type-2 fuzzy sets. To reduce the complexity, interval type-2 fuzzy sets can be used. In this paper, an interval type-2 fuzzy membership design method and its application to radial basis function (RBF) neural networks is proposed. Type-1 fuzzy memberships which are computed from the centroid of the interval type-2 fuzzy memberships are incorporated into the RBF neural network The proposed membership assignment is shown to improve the classification performance of the RBF neural network since the uncertainty of pattern data are desirably controlled by interval type-2 fuzzy memberships. Experimental results for several data sets are given.

Kernel approach to possibilistic C‐means clustering

Kernel approaches can improve the performance of conventional clustering or classification algorithms for complex distributed data. This is achieved by using a kernel function, which is defined as the inner product of two values obtained by a transformation function. In doing so, this allows algorithms to operate in a higher dimensional space (i.e., more degrees of freedom for data to be meaningfully partitioned) without having to compute the transformation. As a result, the fuzzy kernel C‐means (FKCM) algorithm, which uses a distance measure between patterns and cluster prototypes based on a kernel function, can obtain more desirable clustering results than fuzzy C‐means (FCM) for not only spherical data but also nonspherical data. However, it can still be sensitive to noise as in the FCM algorithm. In this paper, to improve the drawback of FKCM, we propose a kernel possibilistic C‐means (KPCM) algorithm that applies the kernel approach to the possibilistic C‐means (PCM) algorithm. The method includes a variance updating method for Gaussian kernels for each clustering iteration. Several experimental results show that the proposed algorithm can outperform other algorithms for general data with additive noise.

A convex cluster merging algorithm using support vector machines

In this paper, we propose a fast and reliable distance measure between two convex clusters using support vector machines (SVM). In doing so, the optimal hyperplane obtained by the SVM is used to calculate the minimal distance between the two clusters. As a result, an effective cluster merging algorithm that groups convex clusters resulted from the fuzzy convex clustering (FCC) method in is developed using this optimal distance. Hence, the number of clusters can be further reduced without losing its representation of the data. Several experimental results are given.

Simulation-based optimization of gas condensate wells to mitigate the heavy hydrocarbon condensation through supercritical CO2 injection

Abstract−In gas-condensate reservoirs suffering from condensate banking, the supercritical CO2 injection process is regarded as one of the most effective technical remedies to reduce the liquid formation and achieve higher quality gas production. With proper well configuration and spacing designs, the injected CO2 can decrease the loss of heavy components effectively. The main goal of this study was to minimize the loss of heavy components during CO2 injection by implementing a proper well configuration. The results show that the integration of pressure maintenance and chemical reactions, including reduced viscosity and interfacial tension, improves the C7+ component recovery by 42.9, 49.4, and 49.3% for the base five-spot, inverted five-spot, and line drive cases, respectively. The total recovery is the highest for the line drive pattern with a recovery factor of 72.7%. The results also indicate that there is a critical length maximizing the effect of gas cycling.

A Density Estimation based Fuzzy C-means Algorithm for Image Segmentation

The Fuzzy E-means (FCM) algorithm is a widely used clustering method that incorporates probabilitic memberships. Due to these memberships, it can be sensitive to noise data. In this paper, we propose a new fuzzy C-means clustering algorithm by incorporating the Parzen Window method to include density information of the data. Several experimental results show that our proposed density-based FCM algorithm outperforms conventional FCM especially for data with noise and it is not sensitive to initial cluster centers.

Integrated Analysis of Permeability Reduction Caused by Polymer Retention for Better Understanding Polymer Transport

Polymer retention is one of the most important factors to govern polymer propagation through porous media, determining whether successful or not. The focus of previous studies has been limited to polymer concentration loss caused by the retention; not only change in polymer concentration, but also reduction in reservoir permeability is the main issue for theoretical transport study. Due to the lack of accuracy of Langmuir isotherm describing the polymer retention mechanisms, this study proposes a new type of matching interpretation method to correlate the permeability reduction factors from experiments to permeability. In order to solve the problem of poorly matching results between estimation and observation, use of nonadsorptive constant conditionally selected in matching process was made. Based on the threshold permeability reduction factors, approximate critical permeability can be calculated to which nonadsorptive constant would be applied. Results showed significant improvements in the estimation of permeability reduction for both low and high permeability cores. In addition, effects of permeability reduction on polymer transport in field scale were analyzed using the proposed matching model. Thus, not only does this interpretation method help to evaluate prediction for accurate flow behavior, but also unwanted risk can be evaluated.

A Kernel based Possibilistic Approach for Clustering and Image Segmentation

The fuzzy kernel c-means (FKCM) algorithm, which uses a kernel function, can obtain more desirable clustering results than fuzzy c-means (FCM) for not only spherical data but also non-spherical data. However, it can be sensitive to noise as in the FCM algorithm. In this paper, a kernel function is applied to the possibilistic c-means (PCM) algorithm and is shown to be robust for data with additive noise. Several experimental results show that the proposed kernel possibilistic c-means (KPCM) algorithm out performs the FKCM algorithm for general data with additive noise.