1,720,965 research outputs found
About Retraining Rule in Multi-Expert Intelligent System for Semi-Supervised learning using SVM classifiers
No full textTraining a system for pattern recognition is a task that require a large amount of
labeled data. However, the creation of such training set is often difficult, expensive and time consuming because it requires the efforts of experienced human annotators. On the other hand, unlabeled data may be relatively easy to collect, but there are few ways to use them. Semi-Supervised learning is a useful approach to reduce human labor and improve
accuracy using unlabeled data, together with labeled data.
This paper proposes three methods in order to re-train classifiers in a multi-expert scenario, when new (unknown) data are available. In fact, when a multi-expert system is adopted, the collective behavior of classifiers can be used both for recognition aims and also selection of
the most profitable samples for system re-train. More specifically a misclassified sample for a particular expert can be used to update the expert itself if the collective behavior of the
multi-expert system allows to classify the sample with high confidence.
In addition, this paper provides a comparison between the new approach and those
available in literature for semi-supervised learning using the SVM classifier by taking into account four different combination techniques at abstract and measurement level. The experimental results, that have been obtained using the handwritten digits of the CEDAR
database, demonstrate the effectiveness of the proposed approach
The Similarity Index lower and upper bounds: Theoretical Considerations and Experimental Verification
No full textIn this paper the Similarity Index variability range is investigated. Depending on the recognition rates of abstract-level classifiers, the lower and upper bounds of the of the Similarity Index variability range is theoretically analysed. The experimental tests, carried out in the field of handwritten numeral classification, confirm the theoretical findings
Supervised Learning Strategies in Multi-Classifier Systems
No full textThis paper presents three strategies in order to re-train classifiers in a multi-expert scenario when new labeled data become available. The simplest possibility is the use of the entire new dataset. The second possibility is related
to the consideration that each single classifier is able to
select new patterns starting from those on which it
performs a miss-classification. Finally, the multi expert system behavior can be inspected to select profitable samples. More specifically a misclassified sample, for a particular classifier, is used to update that classifier only if it produces a miss-classification by the ensemble of classifiers. The three approaches are compared under
different conditions on two different state of the art
performing classifiers by considering the CEDAR
(handwritten digit) database. It is shown how results
depend by the amount of the new training samples, as
well as by the specific combination decision schema
Handwritten processing for Pre Diagnosis of Alzheimer disease
No full textBased on neuromuscular transfer function of the handwriting system, in this paper a non invasive pre diagnosis system for Alzheimer disease alert is proposed. It is well known in fact, that writing originates from spike trains produced within the Central Nervous System (CNS) and more specifically, inside the 4-th and the 6-th regions of the Bradman's map and then transmitted through the first and second order axons to the spinal cord to control the muscles involved in the handwriting as the arm, the forearm, the hand and the pen or pencil utilized for the writing. More specifically, in this work is proposed a new method, not invasive, for early diagnosis of degenerative disability, it can be also useful for monitoring activities related to the progression of neuromuscular disease in order to evaluate the changing related also to the efficiency of the therapies used. Benefit can be obtained not only for the medical field but also for the pharmaceutical developments. Specifically in the paper, the results of some experiments have been focused by considering a certain number of persons some of which affect by Alzheimer disease
Evaluating Threshold for Retraining Rule in Semi-Supervised Learning using Multi-Expert System
No full textThe creation of training set, for pattern recognition, is a difficult, expensive and time consuming task because it requires the efforts of experienced human annotators. On the other hand, unlabeled data can be obtained cheaply, but there are few ways to use them. Semi-Supervised learning uses both labeled and unlabeled data for classification task. In this paper we propose to apply semi-supervised learning and three methods in order to re-train individual classifiers in a multi-expert scenario. More specifically, these experiments are focused on acceptance threshold that defines what data are selected in the feedback-based process. Our approach analyzes the entire system so that a misclassified sample, respect to the final decision, by particular expert can be used to update itself if that sample is classified with a confidence greater than a specific threshold. Experimental results, carried out on the CEDAR (handwritten digits) database, show a comparison between our approach and Self-Training and Co-Training algorithms. The SVM classifier and two different combination techniques at measurement level have been used
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
Multi-Classifier System Configuration using Genetic Algorithms
No full textClassifier combination is a powerful paradigm to deal with difficult pattern classification problems. As matter of this fact, multi-classifier systems have been widely adopted in many applications for which very high classification performance is necessary. Notwithstanding, multi-classifier
system design is still an open problem. In fact, complexity of
multi-classifiers systems make the theoretical evaluation of
system performance very difficult and, consequently, also the
design of a multi-classifier system. This paper presents a new
approach for the design of a multi-classifier system. In
particular, the problem of feature selection for a multiclassifier
system is addressed and a genetic algorithm is
proposed for automatic selecting the optimal set of features for
each individual classifier of the multi-classifier system. The
experimental results, carried out in the field of handwritten
digit recognition, demonstrate the effectiveness of the proposed
approach
Learning Strategies for Knowledge-base Updating in Online Signature Verification Systems
No full textUpdating of reference information is a crucial task for automatic signature verification. In fact, signature characteristics vary in time and whatever approach is considered the effectiveness of a signature verification system strongly depends on the extent to which reference information is able to model the changeable characteristics of users’ signatures. This paper addresses the problem of knowledge-base updating in multi-expert signature verification sys-tems and introduces a new strategy which exploits the collective behavior of classifiers to select the most profitable samples for knowledge-base updating. The experimental tests, carried out using the SUSig database, demonstrate the effectiveness of the new strategy
- …
