Classification Of New Data Research Articles

Visualization of Volcanic Rock Geochemical Data and Classification with Artificial Neural Networks

An unsupervised neural network technique, Growing Cell Structures (GCS) was used to visualize geochemical differences between four different island arc volcanic rock types: basalts, andesites, dacites and rhyolites. The output of the method shows the cluster structure of the dataset clearly, and the relevant geochemical patterns and relationships between its variables. The data can be separated into four clusters, each associated with a specific volcanic rock type (basalt, andesite, dacite and rhyolite), according to a unique combination of major element concentrations. Following clustering, performance of the trained GCS network as a classifier of volcanic rock type was tested using two test datasets with major element concentration data for 312 and 496 island arc volcanic rock samples of known volcanic type. Preliminary classification results are promising. In the first test dataset 94% of basalts, 76% of andesites, 83% of dacites and 100% of the rhyolites were classified correctly. Successful classification rates in the second dataset were 100%, 80%, 77%, and 98% respectively. The success of the analysis suggests that neural networks analysis constitutes a useful analytical tool for identification of natural clusters and examination of the relationships between numeric variables in large datasets, and that can be used for automatic classification of new data.

Considerations for Practical Neural Network Application to a Damage Detection Problem

The application of a multilayer perceptron (MLP) neural network to a damage location problem on a GNAT aircraft wing is considered. The problems associated with effective network training and evaluation are discussed, focussing on ensuring good generalisation performance of the network to the classification of new data. Both conventional Maximum Likelihood and Bayesian Evidence based training techniques are considered and a simple thresholding technique is presented to aid in the rejection of poorly regularised network structures. Examples are presented for an artificial simple 2 class problem (drawn from a Gaussian distribution) and a real 9 class problem on the GNAT aircraft wing.

Tissue counter analysis of histologic sections of melanoma: influence of mask size and shape, feature selection, statistical methods and tissue preparation.

Background: Tissue counter analysis is an image analysis tool designed for the detection of structures in complex images at the macroscopic or microscopic scale. As a basic principle, small square or circular measuring masks are randomly placed across the image and image analysis parameters are obtained for each mask. Based on learning sets, statistical classification procedures are generated which facilitate an automated classification of new data sets. Objective: To evaluate the influence of the size and shape of the measuring masks as well as the importance of feature selection, statistical procedures and technical preparation of slides on the performance of tissue counter analysis in microscopic images. As main quality measure of the final classification procedure, the percentage of elements that were correctly classified was used. Study design: H&E‐stained slides of 25 primary cutaneous melanomas were evaluated by tissue counter analysis for the recognition of melanoma elements (section area occupied by tumour cells) in contrast to other tissue elements and background elements. Circular and square measuring masks, various subsets of image analysis features and classification and regression trees compared with linear discriminant analysis as statistical alternatives were used. The percentage of elements that were correctly classified by the various classification procedures was assessed. In order to evaluate the applicability to slides obtained from different laboratories, the best procedure was automatically applied in a test set of another 50 cases of primary melanoma derived from the same laboratory as the learning set and two test sets of 20 cases each derived from two different laboratories, and the measurements of melanoma area in these cases were compared with conventional assessment of vertical tumour thickness. Results: Square measuring masks were slightly superior to circular masks, and larger masks (64 or 128 pixels in diameter) were superior to smaller masks (8 to 32 pixels in diameter). As far as the subsets of image analysis features were concerned, colour features were superior to densitometric and Haralick texture features. Statistical moments of the grey level distribution were of least significance. CART (classification and regression tree) analysis turned out to be superior to linear discriminant analysis. In the best setting, 95% of melanoma tissue elements were correctly recognized. Automated measurement of melanoma area in the independent test sets yielded a correlation of r=0.846 with vertical tumour thickness (p < 0.001), similar to the relationship reported for manual measurements. The test sets obtained from different laboratories yielded comparable results. Conclusions: Large, square measuring masks, colour features and CART analysis provide a useful setting for the automated measurement of melanoma tissue in tissue counter analysis, which can also be used for slides derived from different laboratories.

Recognition-by-parts: a computational approach to human learning and generalization of shapes.

In this paper human pattern recognition is modeled in terms of how human observers learn to describe patterns in terms of their perceived parts, their unary (part) and binary (relational) attributes and the way in which such attribute states "evidence' different classes of shapes. This approach, originally developed in the area of computer vision, is concerned with algorithms which enable the learning of shape descriptions from examples and the classification of new data (generalization) efficiently and accurately. An object in such an "evidence-based' system is represented by a set of rules, where each rule provides a certain amount of evidence for each object class in the database. The accumulated class evidence over all activated rules can then be used to determine the classification probability. We have examined how well this model reflects human perception by training observers to classify compound Gabor patterns and then testing them with versions of such patterns which were segmented (gray-level transformed) versions of the original training set. If the observers were to construct rules to define each pattern class in terms of perceived parts and their relations, then it should be expected that classification performance would generalize to these new patterns from the original set. Results confirm this hypothesis and the specific feature extraction, learning and rule generation model used to predict performance.

A COMPARISON OF TWO NEW TECHNIQUES FOR CONCEPTUAL CLUSTERING

Clustering, pattern recognition, and classification are important components of many artificial intelligence systems, especially those designed to classify new observations. Often, one has access to a history of old observations that have been previously classified and, under these circumstances, the old data may be used as a training set from which one may obtain rules for the subsequent classification of new data. The techniques used to obtain these rules may be traditional statistical methods or modern computer-intensive techniques. Sometimes, however, the history of old observations has not been previously classified. Under these circumstances, the analyst simply wishes to uncover structure in the data and ascertain whether the structure is apparent or real. When the analyst is searching for clusters, statistical clustering methodologies are often used. Although effective at locating clusters, such approaches leave the interpretation of the clusters as a task for the human analyst. A relatively new class of “conceptual” clustering techniques have emerged from the discipline of machine learning. These techniques attempt to both locate and explain clusters among the data. In this way, the explanations of cluster membership may be used to construct rules for the subsequent classification of new data. The generation of interpretable rules, whether by the use of classification algorithms or conceptual clustering algorithms is of considerable importance in reducing the knowledge acquisition “bottleneck” that often impedes progress towards the building of rule-based systems. In this paper, two new techniques for conceptual clustering are introduced and compared.