Narrow Results

An innovative approach for automated plant disease identification has been proposed in this study. The main contribution of this study is the introduction of a bipartite graph-based clustering technique that has been used for image segmentation, a feature extraction methodology using Self-Organizing Map (SOM), and a ray tracing method. Numerous research works have already been done in this area with their respective merits and demerits. But this bipartite graph-based clustering for image segmentation and feature extraction using SOM and ray tracing techniques has not been used in any of these studies as far as we are aware. The core idea behind this clustering technique is to represent similar spatial data points using a bipartite graph and then singular value decomposition has been used on that graph for clustering. It is common to use SOM for clustering. However, in this study, SOM has been used for feature extraction. First, a spatial dependency matrix based on the pixel value of the gray image has been constructed using SOM. Then some statistical features have been computed from this matrix. Using the ray tracing method, the length of the most extended cluster, i.e. the length of the most extended disease-affected patch, and the distribution of the clusters in the image have been computed. The accuracy of our model has been greatly enhanced using these features only. It has been experimented on disease-affected grape leaf images taken from the Plant Village Dataset. This model outperforms state-of-the-art models, which is shown in the experimental findingsresults section. Not only that, the proposed features produce better accuracy rather than using some existing features. This comparison has also been shown in the experimental findings. The result has been validated using K-fold cross-validation. Last, but not the least, these features produce good accuracy using different classifiers also.

The Self-Organizing Map (SOM) is an efficient tool for visualizing high-dimensional data. In this paper, an intuitive and effective SOM projection method is proposed for mapping high-dimensional data onto the two-dimensional grid structure with a growing self-organizing mechanism. In the learning phase, a growing SOM is trained and the growing cell structure is used as the baseline framework. In the ordination phase, the new projection method is used to map the input vector so that the input data is mapped to the structure of the SOM without having to plot the weight values, resulting in easy visualization of the data. The projection method is demonstrated on four different data sets, including a 118 patent data set and a 399 checical abstract data set related to polymer cements, with promising results and a significantly reduced network size.

Practical data analysis often encounters data sets with both relevant and useless variables. Supervised variable selection is the task of selecting the relevant variables based on some predefined criterion. We propose a robust method for this task. The user manually selects a set of target variables and trains a Self-Organizing Map with these data. This sets a criterion to variable selection and is an illustrative description of the user's problem, even for multivariate target data. The user also defines another set of variables that are potentially related to the problem. Our method returns a subset of these variables, which best corresponds to the description provided by the Self-Organizing Map and, thus, agrees with the user's understanding about the problem. The method is conceptually simple and, based on experiments, allows an accessible approach to supervised variable selection.

The use of self-organizing maps to analyze data often depends on finding effective methods to visualize the SOM's structure. In this paper we propose a new way to perform that visualization using a variant of Andrews' Curves. Also we show that the interaction between these two methods allows us to find sub-clusters within identified clusters. Perhaps more importantly, using the SOM to pre-process data by identifying gross features enables us to use Andrews' Curves on data sets which would have previously been too large for the methodology. Finally we show how a three way interaction between the human user and these two methods can be a valuable exploratory data analysis tool.

The self-organizing map (SOM), as a kind of unsupervised neural network, has been used for both static data management and dynamic data analysis. To further exploit its search abilities, in this paper we propose an SOM-based algorithm (SOMS) for optimization problems involving both static and dynamic functions. Furthermore, a new SOM weight updating rule is proposed to enhance the learning efficiency; this may dynamically adjust the neighborhood function for the SOM in learning system parameters. As a demonstration, the proposed SOMS is applied to function optimization and also dynamic trajectory prediction, and its performance compared with that of the genetic algorithm (GA) due to the similar ways both methods conduct searches.

We present the clustering properties of amino acids, which are building blocks of proteins, according to their physico-chemical characters. To classify the 20 kinds of amino acids, we employ a Self-Organizing Map (SOM) analysis for the Miyazawa-Jernigan (MJ) pairwise-contact matrix, the Environment-dependent One-body energy Parameters (EOP) and the one-body energy parameters incorporating the Ramachandran angle information (EOPR) over the EOP in proteins. We provide the new result of the SOM clustering for amino acids based on the EOPR and compare that with those from the MJ and the EOP matrix. All three kinds of energy parameters capture the leading role played by the hydrophobicity and the hydrophilicity of amino acids in protein folding. Our SOM analysis generally illustrates that both the EOP and the EOPR can provide the collective clustering of amino acids by the side chain characteristics and the secondary structure information. However, EOP is better at classifying amino acids according to their side chain characteristics whereas EOPR is better with secondary structure. We show that the EOP and the EOPR matrix manifests more detailed physico-chemical classification of amino acids than those from the MJ matrix, which does not contain a local environmental information of amino acids in the protein structures.

Community structure is one of the important features of complex networks. Researchers have derived a number of algorithms for detecting communities, some of them suffer from high complexity or need some prior knowledge, such as the size of community or number of communities. For some of them, the quality of the detected community structure cannot be guaranteed, even the results of some of them are nondeterministic. In this paper, we propose a Self-Organizing Map (SOM)-based method for detecting community structure from networks. We first preprocess the network by removing some nodes and their associated edges which have little contribution to the formation of communities, then we construct the extended attribute matrix from the preprocessed network, next we embed the detecting procedure in the training of SOM on the attribute matrix to acquire the initial community structure, and finally, we handle those removed nodes by inserting each of them into the community to which its only neighbor belongs, and fine-tune the initial community structure by merging some of the initial communities to improve the quality of the final result. The performance of the proposed method is evaluated on a variety of artificial networks and real-world networks, and experimental results show that our method takes full advantage of SOM model, it can automatically determine the number of communities embedded in the network, the quality of the detected community structure is steadily promising and superior to those of other comparison algorithms.

A Monte Carlo algorithm for partitioning graphs is presented. The algorithm is based on the self-organizing map, an unsupervised, competitive neural network. A mean-field analysis suggests that the complexity of the algorithm is at most is on the order of O(n³/|E|), where n is the number of vertices of the graph, and |E| the number of edges. This prediction is tested on a class of random graphs. Scaling laws that deviate from the mean-field prediction are obtained. Although the origin of these scaling laws is unclear, their consequences are discussed.

Close to one half of the LHC events are expected to be due to elastic or inelastic diffractive scattering. Still, predictions based on extrapolations of experimental data at lower energies differ by large factors in estimating the relative rate of diffractive event categories at the LHC energies. By identifying diffractive events, detailed studies on proton structure can be carried out.

The combined forward physics objects: rapidity gaps, forward multiplicity and transverse energy flows can be used to efficiently classify proton–proton collisions. Data samples recorded by the forward detectors, with a simple extension, will allow first estimates of the single diffractive (SD), double diffractive (DD), central diffractive (CD), and nondiffractive (ND) cross-sections. The approach, which uses the measurement of inelastic activity in forward and central detector systems, is complementary to the detection and measurement of leading beam-like protons.

In this investigation, three different multivariate analysis approaches are assessed in classifying forward physics processes at the LHC. It is shown that with gene expression programming, neural networks and support vector machines, diffraction can be efficiently identified within a large sample of simulated proton–proton scattering events. The event characteristics are visualized by using the self-organizing map algorithm.

The goal of this work was to develop an improved defect detection scheme for high-speed real-time web surface inspection. This goal was realized by splitting the task into two independent parts: feature extraction and segmentation. Both parts were implemented using efficient algorithms which were implemented in hardware that is suitable and fast enough to be included in a working web inspection system. The proposed scheme is based on some derived texture features and a new self-organizing map variant, the statistical self-organizing map. These techniques offer several improvements over the gray-level thresholding techniques that have been traditionally used in commercial web inspection systems.

We have developed methods for the classification of textures with multidimensional co-occurrence histograms. Gray levels of several pixels with a given spatial arrangement are first compressed linearly and the resulting multidimensional vectors are quantized using the self-organizing map. Histograms of quantized vectors are classified by matching them with precomputed texture model histograms.

In the present study, a multiple resolution technique in linear compression of pixel values is evaluated. The multiple resolution linear compression was made with a local wavelet transform. The vectors were quantized with the tree-structured variant of the self-organizing map. In the tree-structured self-organizing map, the quantization error is reduced, in comparison to the traditional tree-structured codebook, by limited lateral searches in topologically-ordered neighborhoods. The performance of multiresolution texture histograms was compared with single-resolution histograms. The histogram method was compared with three well-established methods: co-occurrence matrices, Gaussian Markov random fields, and multiresolution Gabor energies. The results for a set of natural textures showed that the performance of single- and multiresolution texture histograms was similar. Thus, the benefit of multiresolution analysis was overridden by the multidimensionality of our texture models. Our method gave significantly higher classification accuracies than the three other methods.

This work describes a novel technique for automating the post-fabrication circuit tuning process. A training set that characterizes the behavior of the circuit under test is first constructed. The data in this set consists of input measurement vectors with no output attributes, and is clustered via unsupervised learning algorithm in order to explore its underlying structure and correlations. The generated clusters are labeled and utilized in circuit tuning by calculating the value(s) of the tuning parameter(s). Three prominent and fundamentally different unsupervised learning algorithms, namely, the self-organizing map, the Gaussian mixture model, and the fuzzy C-means algorithm are employed and their performance is compared. The experimental results demonstrate that the proposed technique provides a robust and efficient circuit tuning approach.

In this paper, the authors propose the user evaluation support methods for IGA (Interactive Genetic Algorithm)-based 3D room layout generation. Although IGA is useful for applying the user preferences to the solution, IGA-based systems force users to evaluate a lot of individuals, causing the fatigue of operators. Therefore, we adopted the following methods in order to cope with the problem:

(a) Prediction of user preferences with a conjoint analysis.

(b) Individuals clustering with a self-organizing map.

To make a 3D scene such as a furniture layout, the IGA-based 3D room layout system is very helpful because it requires easy 3D graphics and manipulation only. The authors have been developing a web-based system that optimizes 3D room layouts based on user preferences. Finally, we tested the usefulness of our system with the demonstration, and had positive outcomes.

Recently kernel methods in support vector machines have widely been used in machine learning algorithms to obtain nonlinear models. Clustering is an unsupervised learning method which divides whole data set into subgroups, and popular clustering algorithms such as c-means are employing kernel methods. Other kernel-based clustering algorithms have been inspired from kernel c-means. However, the formulation of kernel c-means has a high computational complexity. This paper gives an alternative formulation of kernel-based clustering algorithms derived from competitive learning clustering. This new formulation obviously uses sequential updating or on-line learning to avoid high computational complexity. We apply kernel methods to related algorithms: learning vector quantization and self-organizing map. We moreover consider kernel methods for sequential c-means and its fuzzy version by the proposed formulation.

Social bookmarking Websites are popular nowadays for they provide platforms that are easy and clear to browse and organize Web pages. Users can add tags on Web pages to allow easy comprehension and retrieval of Web pages. However, tag spams could also be added to promote the opportunity of being referenced of a Web page, which is troublesome to users for accessing uninterested Web pages. In this work, we proposed a scheme to automatically detect such tag spams using a proposed text mining approach based on self-organizing map (SOM) model. We used SOM to find the associations among Web pages as well as tags. Such associations were then used to discover the relationships between Web pages and tags. Tag spams can then be detected according to such relationships. Experiments were conducted on a set of Web pages collected from a social bookmarking site and obtained promising result.

Self-organizing map (SOM) has been used in protein folding prediction when the HP model is employed. The existing work uses a square-like shape lattice with l = m×n points to represent the optimal compact structure of a sequence of l amino acids. In this paper, a general l′-size sequence of amino acids is self-organized in a two dimensional lattice with l (> l′) points. The obtained minimum configuration then has a flexible shape, in contrast to the compact structure limited in the lattice. To fulfil this extension, a new self-organizing map (SOM) technique is proposed to deal with the difficulty of the unsymmetric input and output spaces. New competition rules in the training phase are introduced and a local search method is applied to overcome the multi-mapping phenomena. Several HP benchmark examples with up to 36 amino acids are tested to verify the effectiveness of the proposed approach in this paper.

We propose a hand-eye coordination system for a humanoid robot that supports bimanual reaching. The system combines endpoint closed-loop and open-loop visual servo control. The closed-loop component moves the eyes, head, arms, and torso, based on the position of the target and the robot's hands, as seen by the robot's head-mounted cameras. The open-loop component uses a motor-motor mapping that is learnt online to support movement when visual cues are not available.

In this paper, a feedback self-organizing map (FSOM), which is an extension of the self-organizing map (SOM) by employing feedback loops, is proposed. The SOM consists of an input layer and a competitive layer, and the input vectors applied to the input layer are mapped to the competitive layer keeping their spatial features. In order to embed the temporal information to the SOM, feedback loops from the competitive layer to the input layer are employed. The winner unit in the competitive layer is not assigned by only current input vector but also past winner units, thus the temporal information can be embedded. The effectiveness and validity of the proposed FSOM are verified by applying it to a spatio-temporal pattern classification.

Spatio-temporal connectionist networks comprise an important class of neural models that can deal with patterns distributed in both time and space. In this article, we present new models of self-organizing maps for sequence clustering and classification. We have introduced the temporal dynamics in these maps and we have proposed several new models based on covariance matrices computation. In the first models, the inputs are modeled using its associated covariance matrix. These models, used in speaker recognition, do not take into account the order of the vectors in the sequence. To overcome this drawback, we have proposed new models, which introduce the temporal dynamics in the covariance matrix associated to the input sequences. In order to obtain a network that can learn new knowledge without forgetting the previous learned ones, we have introduced the plasticity and stability properties into one proposed temporal model using the adaptive resonance theory paradigm.

This paper introduces a probabilistic self-organizing map for topographic clustering, analysis and visualization of multivariate binary data or categorical data using binary coding. We propose a probabilistic formalism dedicated to binary data in which cells are represented by a Bernoulli distribution. Each cell is characterized by a prototype with the same binary coding as used in the data space and the probability of being different from this prototype. The learning algorithm, Bernoulli on self-organizing map, that we propose is an application of the EM standard algorithm. We illustrate the power of this method with six data sets taken from a public data set repository. The results show a good quality of the topological ordering and homogenous clustering.