Fuzzy Self-organizing Maps Research Articles

A Novel Approach Using Fuzzy Self-Organizing Maps for Detecting Software Faults

A Novel Approach Using Fuzzy Self-Organizing Maps for Detecting Software Faults

Abstract P6-05-04: Stratification of breast carcinomas with double-equivocal HER2 status

Abstract BACKGROUND AND RATIONALE: Immunohistochemistry (IHC) and in situ hybridization (ISH) represent the cornerstone of HER2 evaluation in breast cancer (BC). We have demonstrated that up to 13% of BCs with equivocal HER2 expression (score 2+ in IHC) harbor an equivocal HER2 gene status (HER2/CEP17 <2 and mean HER2 copy number between 4 and 6), leading to the definition of the double-equivocal category. When we assessed HER2 gene levels in these cases by a PCR-based method we observed copy gain in 25%. When we tested HER2 protein levels by an in situ quantitative assay, HER2 levels ranged from those of score 0/ISH- to those observed in score 2+/ISH+ BCs. These data suggest that, rather than exploring alternative methods to assess HER2 status, a complementary functional approach may be beneficial. Our aim was to stratify double-equivocal BCs using global transcriptomics. METHODS & RESULTS: We retrieved a series of 27 formalin fixed paraffin embedded double-equivocal BCs and two control groups matched for oestrogen receptor and histological grade: 22 HER2- (IHC score 0/ISH-) and 22 HER2+ (IHC score 3+/ISH+) BCs. RNA was extracted following microdissection to enrich for tumor cell content >80% and subjected to whole-Genome DASL (cDNA-mediated Annealing, Selection, extension and Ligation; Illumina). We first identified genes with differential expression in HER2+ versus HER2- BCs based on T-test significance (p<0.01) and on mean gene expression variations higher than +/- 2-fold. To best characterize the signature we performed a cluster analysis with the GEDAS software using the "Fuzzy Self-organizing Maps" algorithm and the cosenic distance to generate the clusters. The classifier (24 genes) tested on double-equivocal cases led to a separation in "HER2+ like" and "HER2- like" BCs. More in details, three main clusters emerged, showing a significantly different distribution of cases with distinct HER2 status (p<0.0001, Chi square). Cluster A included all HER2+ and 5 double-equivocal cases and was associated to high expression of genes that pertain to the HER2 amplicon. Cluster B, composed of 9 HER2- and 12 double-equivocal cases, showed low expression of HER2 and HER2 amplicon-related genes and high levels of expression of TPRG1, NOVA1, AGTR1, SEZ6L, MAPT, GSTM1, SORCS1, DSCR6, NPY1R genes. Cluster C, formed by 13 HER2- and 10 double-equivocal cases, showed low expression of HER2 and HER2 amplicon-related genes but a non-homogeneous expression of the other genes of the classifier. The expression of best performing genes of the classifier (TPRG1, NOVA1, AGTR1) was investigated in The Cancer Genome Atlas (TCGA) dataset of breast cancer (Breast Invasive Carcinoma, TCGA provisional from cbioportal.org) and a striking mutually exclusive pattern with HER2 expression was observed. CONCLUSION: By resolving HER2 double-equivocal BCs into HER2 likely-positive or likely-negative, this approach may pave the way to an informed therapeutic decision in this controversial category of patients. Citation Format: Marchio C, Trisolini E, Maletta F, Annaratone L, Scalzo MS, Mascali D, Verdun di Cantogno L, Medico E, Sapino A. Stratification of breast carcinomas with double-equivocal HER2 status. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P6-05-04.

A Further Study on Mining DNA Motifs Using Fuzzy Self-Organizing Maps.

Self-organizing map (SOM)-based motif mining, despite being a promising approach for problem solving, mostly fails to offer a consistent interpretation of clusters with respect to the mixed composition of signal and noise in the nodes. The main reason behind this shortcoming comes from the similarity metrics used in data assignment, specially designed with the biological interpretation for this domain, which are not meant to consider the inevitable noise mixture in the clusters. This limits the explicability of the majority of clusters that are supposedly noise dominated, degrading the overall system clarity in motif discovery. This paper aims to improve the explicability aspect of learning process by introducing a composite similarity function (CSF) that is specially designed for the k -mer-to-cluster similarity measure with respect to the degree of motif properties and embedded noise in the cluster. Our proposed motif finding algorithm in this paper is built on our previous work robust elicitation algorithms for discovering (READ) [1] and termed READ Deoxyribonucleic acid motifs using CSFs (READ(csf)), which performs slightly better than READ and shows some remarkable improvements over SOM-based SOMBRERO and SOMEA tools in terms of F-measure on the testing data sets. A real data set containing multiple motifs is used to explore the potential of the READ(csf) for more challenging biological data mining tasks. Visual comparisons with the verified logos extracted from JASPAR database demonstrate that our algorithm is promising to discover multiple motifs simultaneously.

Relational Fuzzy Self-Organizing Maps for Cluster Visualization and Summarization

The notion of Best-Matching Unit (BMU) in the proposed Fuzzy Relational Self-Organizing (FRSOM) algorithm is replaced by a membership function where every neuron has a certain degree of matching to an input object. The FRSOM is an extension of the relational self-organizing map. In the proposed FRSOM we incorporate a monotonically increasing fuzzifier and a monotonically decreasing neighborhood kernel. Initially, FRSOM assigns winning neurons. However, as time progresses adjacent neurons begin communicating and sharing information about the stimulus received. The amount of information being shared at a given time is governed by the fuzzifier and the number of neurons sharing information is controlled by the neighborhood kernel. Additionally, in this paper we show that FRSOM is the relational dual of Fuzzy Batch SOM (FBSOM) followed by experimental results comparing both FBSOM and FRSOM on synthetic datasets. Then we will demonstrate the visualization and summarization capabilities of FRSOM on two real relational datasets, Gene Ontology and a patient data consisting of Activity of Daily Living score trajectories.

Synthetic Aperture Radar Image Clustering with Curvelet Subband Gauss Distribution Parameters

Curvelet transform is a multidirectional multiscale transform that enables sparse representations for signals. Curvelet-based feature extraction for Synthetic Aperture Radar (SAR) naturally enables utilizing spatial locality; the use of curvelet-based feature extraction is a novel method for SAR clustering. The implemented method is based on curvelet subband Gaussian distribution parameter estimation and cascading these estimated values. The implemented method is compared against original data, polarimetric decomposition features and speckle noise reduced data with use of k-means, fuzzy c-means, spatial fuzzy c-means and self-organizing maps clustering methods. Experimental results show that the curvelet subband Gaussian distribution parameter estimation method with use of self-organizing maps has the best results among other feature extraction-clustering performances, with up to 94.94% overall clustering accuracies. The results also suggest that the implemented method is robust against speckle noise.

A robust elicitation algorithm for discovering DNA motifs using fuzzy self-organizing maps.

It is important to identify DNA motifs in promoter regions to understand the mechanism of gene regulation. Computational approaches for finding DNA motifs are well recognized as useful tools to biologists, which greatly help in saving experimental time and cost in wet laboratories. Self-organizing maps (SOMs), as a powerful clustering tool, have demonstrated good potential for problem solving. However, the current SOM-based motif discovery algorithms unfairly treat data samples lying around the cluster boundaries by assigning them to one of the nodes, which may result in unreliable system performance. This paper aims to develop a robust framework for discovering DNA motifs, where fuzzy SOMs, with an integration of fuzzy c-means membership functions and a standard batch-learning scheme, are employed to extract putative motifs with varying length in a recursive manner. Experimental results on eight real datasets show that our proposed algorithm outperforms the other searching tools such as SOMBRERO, SOMEA, MEME, AlignACE, and WEEDER in terms of the F-measure and algorithm reliability. It is observed that a remarkable 24.6% improvement can be achieved compared to the state-of-the-art SOMBRERO. Furthermore, our algorithm can produce a 20% and 6.6% improvement over SOMBRERO and SOMEA, respectively, in finding multiple motifs on five artificial datasets.

Selected Papers from SCIS & ISIS 2010 – No.2

This second special issue from the Joint 5th International Conference on Soft Computing and Intelligent Systems and the 11th International Symposium on Advanced Intelligent Systems (SCIS & ISIS) 2010 presents 11 progressive papers covering topics from Fuzzy c-Means, clustering, evolutionary algorithms, and self-organizing maps to optimization, spiral phenomena, matroids, multirobot systems, and reinforcement learning. This wide range of research interests reflects current trends in the SCIS & ISIS conference series. As such, they demonstrate how near-future research interests are gradually moving toward exciting new directions. We thank the authors and reviewers for taking time out from their busy schedules to make this publication possible. We are especially grateful to Prof. Toshio Fukuda of Nagoya University and to Prof. Kaoru Hirota of the Tokyo Institute of Technology for giving us the opportunity to guest-edit JACIII. Our special thanks go to Ms. Reiko Ohta of Fuji Technology Press Ltd. for her ongoing support in publication.

Fuzzy optimized self-organizing maps and their application to document clustering

In this paper, an approach using fuzzy logic techniques and self-organizing maps (SOM) is presented in order to manage conceptual aspects in document clusters and to reduce the training time. In order to measure the presence degree of a concept in a document, a concept frequency formula is introduced. This formula is based on new fuzzy formulas to calculate the polysemy degree of terms and the synonymy degree between terms. In this approach, new fuzzy improvements such as automatic choice of the topology, heuristic map initialization, a fuzzy similarity measure and a keywords extraction process are used. Some experiments have been carried out in order to compare the proposed system with classic SOM approaches by means of Reuters collection. The system performance has been measured in terms of F-measure and training time. The experimental results show that the proposed approach generates good results with less training time compared to classic SOM techniques.

Selected Papers from SCIS & ISIS 2008 – No.1

Welcome to this special issue of the Journal of Advanced Computational Intelligence and Intelligent Informatics (JACIII). I am pleased to introduce 41 selected papers presented at the 3rd International Conference on Soft Computing and Intelligent Systems (SCIS) and the 7th International Symposium on Advanced Intelligent Systems (ISIS) held on September 17-21, 2008, at Nagoya University in Nagoya, Japan. This conference featured 401 original papers in presentations attended by some 500 participants. SCIS & ISIS is a biennial international joint conference in the field of soft computing and intelligent systems, including branches of research ranging from fuzzy systems, neural networks, and evolutionary computation to multiagent systems, artificial intelligence, and robotics. This current issue presents 20 papers covering most of the conference topics including fuzzy theory, self-organizing maps, robotics, computer vision, and optimization algorithms. I would like to thank the authors and reviewers and SCIS & ISIS 2008 for making this special issue possible. I am also grateful to Prof. Toshio Fukuda, Nagoya University, and Prof. Kaoru Hirota, Tokyo Institute of Technology, the editors-in-chief, and the SCIS & ISIS 2008 conference staff for inviting me to guest-edit this Journal.

FLAME, a novel fuzzy clustering method for the analysis of DNA microarray data.

BackgroundData clustering analysis has been extensively applied to extract information from gene expression profiles obtained with DNA microarrays. To this aim, existing clustering approaches, mainly developed in computer science, have been adapted to microarray data analysis. However, previous studies revealed that microarray datasets have very diverse structures, some of which may not be correctly captured by current clustering methods. We therefore approached the problem from a new starting point, and developed a clustering algorithm designed to capture dataset-specific structures at the beginning of the process.ResultsThe clustering algorithm is named Fuzzy clustering by Local Approximation of MEmbership (FLAME). Distinctive elements of FLAME are: (i) definition of the neighborhood of each object (gene or sample) and identification of objects with "archetypal" features named Cluster Supporting Objects, around which to construct the clusters; (ii) assignment to each object of a fuzzy membership vector approximated from the memberships of its neighboring objects, by an iterative converging process in which membership spreads from the Cluster Supporting Objects through their neighbors. Comparative analysis with K-means, hierarchical, fuzzy C-means and fuzzy self-organizing maps (SOM) showed that data partitions generated by FLAME are not superimposable to those of other methods and, although different types of datasets are better partitioned by different algorithms, FLAME displays the best overall performance. FLAME is implemented, together with all the above-mentioned algorithms, in a C++ software with graphical interface for Linux and Windows, capable of handling very large datasets, named Gene Expression Data Analysis Studio (GEDAS), freely available under GNU General Public License.ConclusionThe FLAME algorithm has intrinsic advantages, such as the ability to capture non-linear relationships and non-globular clusters, the automated definition of the number of clusters, and the identification of cluster outliers, i.e. genes that are not assigned to any cluster. As a result, clusters are more internally homogeneous and more diverse from each other, and provide better partitioning of biological functions. The clustering algorithm can be easily extended to applications different from gene expression analysis.

Part family formation through fuzzy ART2 neural network

In order to overcome some unavoidable factors, like shift of the part, that influence the crisp neural networks' recognition, the present study is dedicated in developing a novel fuzzy neural network (FNN), which integrates both the fuzzy set theory and adaptive resonance theory 2 (ART2) neural network for grouping the parts into several families based on the image captured from the vision sensor. The proposed network posses the fuzzy inputs as well as the fuzzy weights. The model evaluation results showed that the proposed fuzzy neural network is able to provide more accurate results compared to the fuzzy self-organizing feature maps (SOM) neural network [R.J. Kuo, S.S. Chi, P.W. Teng, Generalized part family formation through fuzzy self-organizing feature map neural network, International Journal of Computers in Industrial Engineering, 40 (2001b) 79–100] and fuzzy c-means algorithm.

Clustering properties of fuzzy Kohonen's self-organizing feature maps

A new clustering algorithm called fuzzy self-organizing feature maps is introduced. It can process not only the exact digital inputs, but also the inexact or fuzzy non-digital inputs, such as natural language inputs. Simulation results show that the new algorithm is superior to original Kohonen's algorithm in clustering performance and learning rate.