1,720,963 research outputs found
A new Cursive Basic Word Database for Bank-check Processing Systems
No full textIn this paper is presented a new database for
handwritten cursive basic words recognition. The database
is devoted to research on bank-check processing. In fact,
for security reasons, the banks rarely allow the treatment
of checks handled by them. On the other hand, the reasons
for the English basic words choice lie in the fact that
English language is the world most commonly used for
bank-check drawing. The database realised includes a
considerable number of instances of basic words. Pattern
images are stored using a standard image format that will
be available to all researchers by Internet. The importance
of this work lies in the fact that the database is queried by
the network, giving the possibility to grow with the
contribution of others researchers. The tagging has been
generated by using the XML language that allows
recovering also information on the writers. Furthermore,
the information handled not only could allow the semantic
recognition of the specimen but also the research
development on the author identity of the manuscript. The
database is an open one to be increased with the
contribution of all others researchers in the world
A Novel Technique for Handwritten Digit Classification using Genetic Clustering
No full textThe aim of this paper is to introduce a novel technique for handwritten digit recognition based on genetic clustering. Cluster design is proposed as a two-step process. The first step is focused on generating cluster solutions, while the second one involves the construction of the best cluster solution starting from a set of suitable candidates. An approach for achieving these goals is presented. Clustering is considered as an optimization problem in which the objective function to be minimized is the cost function associated to the classification. A genetic algorithm is used to determine the best cluster centers to reduce classification time, without greatly affecting the accuracy. The classification task is performed by k-nearest neighbor classifier. It has also been developed a new feature and a distance measure based on the Sokal-Michener dissimilarity measure to describe and compare handwritten numerals. This technique has been evaluated through experimental testing on MNIST dataset and its effectiveness has been proved
ADAPTIVE ZONING DESIGN BY SUPERVISED LEARNING USING MULTI-OBJECTIVE OPTIMIZATION
No full textZoning is a widespread feature extraction technique for handwritten digit recognition, since it is able to handle handwritten pattern variability. Static techniques for zoning design have recently been superseded by adaptive techniques, in which zoning design is considered as the result of an optimization procedure. This paper presents a new learning strategy to optimal zoning design using multi-objective genetic algorithm. More precisely, the nondominant sorting genetic algorithm II (NSGA II) has been applied to define, in a single process, both the optimal number of zones and the optimal zones for the Voronoi-based zoning method. The experimental tests, carried out in the field of handwritten digit recognition, show the effectiveness of this new approach with respect to traditional dynamic approaches for zoning design, based on single-objective optimization techniques
Genetic Algorithm Based Clustering Approach for Improving Off-line Handwritten Digit Classification
No full textIn this paper a new clustering technique for improving off-line
handwritten digit recognition is introduced. Clustering design is
approached as an optimization problem in which the objective
function to be minimized is the cost function associated to the
classification, that is here performed by the k-nearest neighbor (k-
NN) classifier based on the Sokal and Michener dissimilarity
measure. For this purpose, a genetic algorithm is used to determine
the best cluster centers to reduce classification time, without
suffering a great loss in accuracy. In addition, an effective strategy
for generating the initial-population of the genetic algorithm is also
presented. The experimental tests carried out using the MNIST
database show the effectiveness of this method
Handwritten Digit Recognition by Multi-Objective Optimization of Zoning Methods
No full textThis paper addresses the use of multi-objective
optimization techniques for optimal zoning design in the context of handwritten digit recognition. More precisely, the Non-dominant Sorting Genetic Algorithm II (NSGA II) has been considered for the optimization of Voronoi-based zoning
methods. In this case both the number of zones and the zone position and shape are optimized in a unique genetic procedure. The experimental results point out the usefulness
of multi-objective genetic algorithms for achieving effective zoning topologies for handwritten digit recognition
A New Adaptive Zoning Technique for Handwritten Digit Recognition
No full textIn this paper we present a new adaptive zoning technique based on Voronoi tessellation for the task of handwritten digit recognition. This technique extracts features according to an optimal zoning distribution, obtained by an evolutionary-strategy based search. Several experiments have been conducted on the MNIST and the USPS datasets to investigate the proposed approach. Comparisons with regular square zoning reveal that the presented zoning strategy achieves better results for any type of SVM classifier. Furthermore, the proposed zoning method shows that the combination of the adaptive zoning strategy with the Voronoi topology leads to find a distribution of zones able to improve accuracy significantly. As a matter of fact reached accuracies are close to the best algorithms
Learning Iterative Strategies in Multi-Expert Systems using SVMs for Digit Recognition
No full textThis paper presents three different learning iterative strategies, in a
multi-expert system. In first strategy entire new dataset is used. In second strategy
each single classifier selects new samples starting from those on which it
performs a misclassification. Finally, the collective behavior of classifiers is
studied to select the most profitable samples for knowledge base updating. The
experimental results provide a comparison of three approaches under different
operating conditions and feedback process. A classifier SVM and four different
combination techniques were used by considering the CEDAR (handwritten
digit) database. It is shown how results depend by the iterations on the feedback
process, as well as by the specific combination decision schema and by data
distribution
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
