Search results for "algorithm."

Pattern classification using a new border identification paradigm: The nearest border technique

2015

Abstract There are many paradigms for pattern classification such as the optimal Bayesian, kernel-based methods, inter-class border identification schemes, nearest neighbor methods, nearest centroid methods, among others. As opposed to these, this paper pioneers a new paradigm, which we shall refer to as the nearest border (NB) paradigm. The philosophy for developing such a NB strategy is as follows: given the training data set for each class, we shall attempt to create borders for each individual class. However, unlike the traditional border identification (BI) methods, we do not undertake this by using inter-class criteria; rather, we attempt to obtain the border for a specific class in t…

Efficient learning of regular expressions from good examples

1994

We consider the problem of restoring regular expressions from expressive examples. We define the class of unambiguous regular expressions, the notion of the union number of an expression showing how many union operations can occur directly under any single iteration, and the notion of an expressive example. We present a polynomial time algorithm which tries to restore an unambiguous regular expression from one expressive example. We prove that if the union number of the expression is 0 or 1 and the example is long enough, then the algorithm correctly restores the original expression from one good example. The proof relies on original investigations in theory of covering symbol sequences (wo…

A new paradigm for pattern classification: Nearest Border Techniques

2013

Published version of a chapter in the book: AI 2013: Advances in Artificial Intelligence. Also available from the publisher at: http://dx.doi.org/10.1007/978-3-319-03680-9_44 There are many paradigms for pattern classification. As opposed to these, this paper introduces a paradigm that has not been reported in the literature earlier, which we shall refer to as the Nearest Border (NB) paradigm. The philosophy for developing such a NB strategy is as follows: Given the training data set for each class, we shall first attempt to create borders for each individual class. After that, we advocate that testing is accomplished by assigning the test sample to the class whose border it lies closest to…

On the zeros of Jacobi polynomials

1994

Variability of Classification Results in Data with High Dimensionality and Small Sample Size

2021

The study focuses on the analysis of biological data containing information on the number of genome sequences of intestinal microbiome bacteria before and after antibiotic use. The data have high dimensionality (bacterial taxa) and a small number of records, which is typical of bioinformatics data. Classification models induced on data sets like this usually are not stable and the accuracy metrics have high variance. The aim of the study is to create a preprocessing workflow and a classification model that can perform the most accurate classification of the microbiome into groups before and after the use of antibiotics and lessen the variability of accuracy measures of the classifier. To ev…

Integrated fuzzy classification

2003

Span Programs and Quantum Algorithms for st-Connectivity and Claw Detection

2012

We introduce a span program that decides st-connectivity, and generalize the span program to develop quantum algorithms for several graph problems. First, we give an algorithm for st-connectivity that uses O(n d^{1/2}) quantum queries to the n x n adjacency matrix to decide if vertices s and t are connected, under the promise that they either are connected by a path of length at most d, or are disconnected. We also show that if T is a path, a star with two subdivided legs, or a subdivision of a claw, its presence as a subgraph in the input graph G can be detected with O(n) quantum queries to the adjacency matrix. Under the promise that G either contains T as a subgraph or does not contain T…

Evaluation of enantioselective binding of fluoxetine to human serum albumin by ultrafiltration and CE - Experimental design and quality considerations

2012

Several pharmacokinetic processes are affected by enantioselectivity (ES). At the level of distribution, protein binding (PB) is one of the most important. The enantioselective binding of fluoxetine (FLX) to HSA has been evaluated in this work by ultrafiltration of FLX–HSA mixtures and chiral analysis of unbound fractions by EKC-CD. PB, affinity constants (K) and ES were obtained for both enantiomers of FLX. In order to improve the consistency of the estimations, the evaluation of affinity constants of each enantiomer was performed using two designs, one keeping constant the total concentration of protein and varying the total concentration of the enantiomers, and the other in the opposite …

Maximum weight relaxed cliques and Russian Doll Search revisited

2015

Trukhanov et al. [Trukhanov S, Balasubramaniam C, Balasundaram B, Butenko S (2013) Algorithms for detecting optimal hereditary structures in graphs, with application to clique relaxations. Comp. Opt. and Appl., 56(1), 113–130] used the Russian Doll Search (RDS) principle to effectively find maximum hereditary structures in graphs. Prominent examples of such hereditary structures are cliques and some clique relaxations intensely discussed and studied in network analysis. The effectiveness of the tailored RDS by Trukhanov et al. for s-plex and s-defective clique can be attributed to their cleverly designed incremental verification procedures used to distinguish feasible from infeasible struct…

Efficient unsupervised clustering for spatial bird population analysis along the Loire river

2015

International audience; This paper focuses on application and comparison of Non Linear Dimensionality Reduction (NLDR) methods on natural high dimensional bird communities dataset along the Loire River (France). In this context, biologists usually use the well-known PCA in order to explain the upstream-downstream gradient.Unfortunately this method was unsuccessful on this kind of nonlinear dataset.The goal of this paper is to compare recent NLDR methods coupled with different data transformations in order to find out the best approach. Results show that Multiscale Jensen-Shannon Embedding (Ms JSE) outperform all over methods in this context.