This paper proposes a graph-based evolutionary algorithm called Genetic Network Programming (GNP). Our goal is to develop GNP, which can deal with dynamic environments efficiently and effectively, based on the distinguished expression ability of graph (network) structure. The characteristics GNP are as follows. 1) programs composed number nodes execute simple judgment/processing, these connected by directed links each other. 2) structure enables re-use nodes, thus be very compact. 3) node...

Image segmentation plays an important role in image analysis and computer vision system. Among all techniques, the automatic thresholding methods are widely used because of their advantages simple implement time saving. Otsu method is one frequently various fields. Two-dimensional (2D) behaves well segmenting images low signal-to-noise ratio than one-dimensional (1D). But it gives satisfactory results only when numbers pixels each class close to other. Otherwise, improper results. In this...

Multiagent Systems with Symbiotic Learning and Evolution (Masbiole) has been proposed studied, which is a new methodology of (MAS) based on symbiosis in the ecosystem. Masbiole employs method symbiotic learning evolution where agents can learn or evolve according to their relations toward others, i.e., considering benefits/losses both itself an opponent. As result, escape from Nash Equilibria obtain better performances than conventional MAS consider only own benefits. This paper focuses...

In this paper, an SVM-based approach is proposed for stock market trend prediction. The consists of two parts: feature selection and prediction model. the part, a correlation-based SVM filter applied to rank select good subset financial indexes. And indicators are evaluated based on ranking. model so called quasi-linear predict movement direction in term historical data series by using selected indexes as weighted inputs. with composite kernel function, which approximates nonlinear...

Recently, many methods of evolutionary computation such as genetic algorithm (GA) and programming (GP) have been developed a basic tool for modeling optimizing complex systems. Generally speaking, GA has the genome string structure, while in GP is tree structure. Therefore, suitable constructing complicated programs, which can be applied to real world problems. However, might sometimes difficult search solution because its bloat. A novel method named Genetic Network Programming (GNP), whose...

A method of association rule mining using Genetic Network Programming (GNP) is proposed to improve the performance extraction. The mechanisms can calculate measurements rules directly GNP, and measure significance via chi-squared test. Users define conditions importance flexibly, which include value number attributes in a rule. system evolves itself by an evolutionary obtains candidates genetic operations. Extracted are stored pool all together through generations reflected operators as...

An efficient algorithm for important class association rule mining using genetic network programming (GNP) is proposed. GNP one of the evolutionary optimization techniques, which uses directed graph structures as genes. Instead generating a large number candidate rules, method can obtain sufficient rules classification. The proposed measures significance via chi-squared test. Therefore, all extracted be used classification directly. In addition, suits from dense databases, where many...

Evaluating image captions is very challenging partially due to the fact that there are multiple correct for every single image. Most of existing one-to-one metrics operate by penalizing mismatches between reference and generative caption without considering intrinsic variance ground truth captions. It usually leads over-penalization thus a bad correlation human judgment. Recently, latest metric BERTScore can achieve high in system-level tasks while some issues be fixed better performance. In...

To develop an efficient brain-computer interface (BCI) system, electroencephalography (EEG) measures neuronal activities in different brain regions through electrodes. Many EEG-based motor imagery (MI) studies do not make full use of network topology. In this paper, a deep learning framework based on modified graph convolution neural (M-GCN) is proposed, which temporal-frequency processing performed the data S-transform (MST) to improve decoding performance original EEG signals types MI...

Brain-Computer Interface (BCI) is a system provides an alternative communication and control channel between the human brain computer. In Motor Imagery-based (MI) BCI system, Common Spatial Pattern (CSP) frequently used for extracting discriminative patterns from electroencephalogram (EEG). There are many studies have proven that performance of CSP has very important relation with choice operational frequency band. As fact features at different bands contain complementary information...

This paper focuses on imbalanced dataset classification problem by using SVM and oversampling method. Traditional method increases the occurrence of over-lapping between classes, which leads to poor generalization classification. To solve this proposes a combined quasi-linear assembled SMOTE. The is an with kernel function. It realizes approximate nonlinear separation boundary mulit-local linear boundaries interpolation. SMOTE implements considering data distribution information avoids...

This paper proposes a class of quasi-ARMAX models for non-linear systems. Similar to ordinary ARMAX models, the are flexible black-box but they have various linearity properties similar those linear models. A modelling scheme is introduced construct consisting two parts: macro-part and kernel-part. By using Taylor expansion other mathematical transformation techniques, it first constructed as interfaces (macro-parts) that contain some complicated coefficients. MIMO neurofuzzy (kernel-parts)...

This paper presents an adaptive neural network controller (ANNC) that is used to control the speed of a separately excited DC motor driving centrifugal pump load and fed from photovoltaic (PV) generator through DC-DC buck-boost converter. The also track maximum power point (MPP) PV by controlling converter duty ratio. Such kinds controllers must have two objective functions perform these tasks, but in this research function related MPP converted constrained for second making some...

In a standard support vector machine (SVM), the training process has O(n <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ) time and xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> space complexities, where n is size of dataset. Thus, it computationally infeasible for very large datasets. Reducing dataset naturally considered to solve this problem. SVM classifiers depend on only vectors (SVs) that lie close separation boundary....

This paper proposes a so called quasi-linear support vector machine (SVM), which is an SVM with composite kernel. In the model, nonlinear separation hyperplane approximated by multiple local linear models interpolation. Instead of building separately, realizes multi model approach in kernel level. That is, it built exactly same way as single composing A guided partitioning method proposed to obtain partitions for composition function. Experiment results on artificial data and benchmark...

Alternative splicing (AS) is an important post-transcriptional process. It has been suggested that most AS events are subject to tissue-specific regulation. However, the global dynamics of in different tissues poorly explored. To analyse changes multiple tissues, we identified and constructed a comprehensive catalogue within each tissue based on genome-wide RNA-seq reads from ten cucumber. First, found 58% multi-exon genes underwent AS. We further obtained 565 with significantly more...

Air pollution has threaten people's health. It is urgent for the government to strengthen and enhance air monitoring capacity. In this paper, we propose an quality prediction model infer pollutant concentrations, such as CO, NO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> , xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> . The idea design a sophisticate piecewise linear by using gated network. A top k% winner-take-all autoencoder...