This paper investigates the use of evolutionary programming for the search of hypothesis space in visual learning tasks. The general goal of the project is to elaborate human-competitive procedures for pattern discrimination by means of learning based on the training data (set of images). In particular, the topic addressed here is the comparison between the "standard" genetic programming (as defined by Koza [13] and the genetic programming extended by local optimization of solutions, so-called genetic local search. The hypothesis formulated in the paper is that genetic local search provides better solutions (i.e. classifiers with higher predictive accuracy) than the genetic search without that extension. This supposition was positively verified in an extensive comparative experiment of visual learning concerning the recognition of handwritten characters.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Characterization of the time-series analysis is presented, as a data mining tool which facilitates better understanding nature of manufacturing process and permits forecasting of future values of the process parameters or production results on the basis of the past data, recorded in regular intervals. The main methods and problems of the time-series analysis are presented, related to the trend function, evaluation of seasonality and significance of the information contents in the residual values. The authors' research results, related to exemplary production data collected in a foundry with Disamatic molding line (temperature of the molding sand), are presented. It is concluded that a properly performed analysis of time-series can be a useful tool for analysis and predictions of foundry production process.
We propose a method of knowledge reuse for an ensemble of genetic programming-based learners solving a visual learning task. First, we introduce a visual learning method that uses genetic programming individuals to represent hypotheses. Individuals-hypotheses process image representation composed of visual primitives derived from given training images that contain objects to be recognized. The process of recognition is generative, i.e. an individual is suposed to restore the shape of the processed object by drawing its reproduction on a separate canvas. This canonical method is in the following extended with a knowledge reuse mechanism that allows a learner to import genetic material from hypotheses that evolved for other decision classes (object classes). We compare the performance of the extended approach to the basis method on a real-world tasks of handwritten character recognition, and conclude that knowledge reuse leads to significant convergence speedup and reduces the risk of overfitting.
4
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
The forecast of structure and properties of casting is based on results of computer simulation of physical processes which are carried out during the casting processes. For the effective using of simulation system it is necessary to validate mathematica-physical models describing process of casting formation and the creation of local discontinues, witch determinate the casting properties. In the paper the proposition for quantitative validation of VP system using solidification casting defects by information sources of II group (methods of NDT) was introduced. It was named the VP/RT validation (virtual prototyping/radiographic testing validation). Nowadays identification of casting defects noticeable on X-ray images bases on comparison of X-ray image of casting with relates to the ASTM. The results of this comparison are often not conclusive because based on operator's subjective assessment. In the paper the system of quantitative identification of iron casting defects on X-ray images and classification this defects to ASTM class is presented. The methods of pattern recognition and machine learning were applied.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.