Counterpropagation Neural Network Research Articles

Counter-propagation artificial neural network models in read-across predictions of toxicity

Counter-propagation artificial neural network models in read-across predictions of toxicity

Novel Method Proposing Chemical Structures with Desirable Profile of Activities Based on Chemical and Protein Spaces.

Active molecules among numerous chemical structures in a chemical database can be searched easily by statistical prediction of compound-protein interactions. However, constructing a simple prediction model against one protein does not aid drug design, because detecting chemical structures that act similarly against multiple proteins is necessary for preventing side effects of the potential drug. To tackle this problem, we propose a new method that visualizes chemical and protein spaces. For simultaneous visualization of both spaces, we employ a counterpropagation neural network (CPNN) and develop a new visualization method named multi-input CPNN (MICPNN). In a case study of the kinase protein family, the MICPNN model predicted accurately the complex relationships between compounds and proteins. The proposed method identified chemical structures with promising activity against kinases. Our proposed method is also applicable to other protein families, such as G-protein coupled receptors, ion channels and transporters.

Image Recognition from Face Feature Descriptor Using Counter Propagation Neural Network

Image Recognition from Face Feature Descriptor Using Counter Propagation Neural Network

CPANNatNIC software for counter-propagation neural network to assist in read-across

BackgroundCPANNatNIC is software for development of counter-propagation artificial neural network models. Besides the interface for training of a new neural network it also provides an interface for visualisation of the results which was developed to aid in interpretation of the results and to use the program as a tool for read-across.ResultsThe work presents the details of the program’s interface. Parts of the interface are presented and how they can be used. The examples provided show how the user can build a new model and view the results of predictions using the interface. Examples are given to show how the software may be used in read-across.ConclusionsCPANNatNIC provides a simple user interface for model development and visualisation. The interface implements options which may simplify read-across procedure. Statistical results show better prediction accuracy of read-across predictions than model predictions where similar compounds could be identified, which indicates the importance of using read-across and usefulness of the program.

Hybridisation of Mel Frequency Cepstral Coefficient and Higher Order Spectral Features for Musical Instruments Classification

Abstract This paper presents the classification of musical instruments using Mel Frequency Cepstral Coefficients (MFCC) and Higher Order Spectral features. MFCC, cepstral, temporal, spectral, and timbral features have been widely used in the task of musical instrument classification. As music sound signal is generated using non-linear dynamics, non-linearity and non-Gaussianity of the musical instruments are important features which have not been considered in the past. In this paper, hybridisation of MFCC and Higher Order Spectral (HOS) based features have been used in the task of musical instrument classification. HOS-based features have been used to provide instrument specific information such as non-Gaussianity and non-linearity of the musical instruments. The extracted features have been presented to Counter Propagation Neural Network (CPNN) to identify the instruments and their family. For experimentation, isolated sounds of 19 musical instruments have been used from McGill University Master Sample (MUMS) sound database. The proposed features show the significant improvement in the classification accuracy of the system.

Modified Chaotic Bat Algorithm Based Counter Propagation Neural Network for Uncertain Nonlinear Discrete Time System

Weight and bias connection are important features of neural networks, which is still challenging for researchers. In this work, we focus on initial weights and bias connection of counter propagation network (CPN) using modified chaotic bat algorithm (MCBA) i.e., MCBA-CPN for uncertain nonlinear systems and compare it with CPN using chaotic bat algorithm (CBA) i.e., CBA-CPN. Chaotic function is used for pulse frequency of bats in MCBA. We have implemented CBA and MCBA, which are based on the consideration of the global solution in the sound intensity adjustment. MCBA-CPN is applied on different uncertain nonlinear systems and Mackey–Glass time series data to test the concert in terms of prediction accuracy. Proposed method is validated through statistical testing like chi-square and [Formula: see text]-test demonstrate that the difference between target and output of proposed method are acceptable. Finally, MCBA-CPN is applied to a real world problem for prediction of milk production data.

An evaluation of regionalization and watershed classification schemes for continuous daily streamflow prediction in ungauged watersheds

Regionalization – the process of transferring hydrological information from gauged to ungauged watersheds – has the potential to perform significantly better if these watersheds are classified in advance. In this study, we demonstrate the benefits of classification by a systematic combination of watershed classification techniques, regionalization methods, and rainfall-runoff models. Basins are first classified, then regionalization methods are applied, for continuous daily streamflow estimation at ungauged watersheds in Ontario, Canada. Nonlinear data-driven methods are used as regionalization and watershed classification schemes to transfer the parameters of two conceptual hydrologic models – namely McMaster University Hydrologiska Byråns Vattenbalansavdelning (MAC-HBV) and Sacramento Soil Moisture Accounting (SAC-SMA) – from gauged to ungauged watersheds. Our results suggest that a certain combination of watershed classification technique, regionalization method and hydrologic model can significantly improve the estimation of continuous streamflow at ungauged basins by improving the accuracy of estimated daily mean, low and peak flows. However, some combinations do not provide a clear improvement when compared to the scenario of unclassified basins. For example, the MAC-HBV model coupled with a counter propagation neural network as a regionalization technique provides a clear improvement in estimated daily mean, low and peak flow when the watersheds are first classified using a nonlinear principal component analysis method. Interestingly, a higher improvement is achieved for low flow as well, which is usually difficult to estimate in ungauged basins.

Neural networks for pattern-based short-term load forecasting: A comparative study

In this work several univariate approaches for short-term load forecasting based on neural networks are proposed and compared. They include: multilayer perceptron, radial basis function neural network, generalized regression neural network, fuzzy counterpropagation neural networks, and self-organizing maps. A common feature of these methods is learning from patterns of the seasonal cycles of load time series. Patterns used as input and output variables simplify the forecasting problem by filtering out a trend and seasonal variations of periods longer than a daily one. Nonstationarity in mean and variance is also eliminated. In the simulation studies using real power system data the neural network forecasting methods were tested and compared with other popular forecasting methods such as ARIMA and exponential smoothing. The best results were achieved for generalized regression neural network and one-neuron perceptron learned locally.

Structural Impairment Detection Using Deep Counter Propagation Neural Networks

Structural systems may deteriorate under normal operating conditions during their service life. In many situations, damage is not easily observed through traditional visual inspection techniques. The need for detecting subtle structural damage and assessing the level of impairment presents opportunities to develop innovative solutions beyond typical inspections. One method for monitoring and evaluating structures is to measure responses and assess the impairment condition of structures using those responses. This paper presents an inverse static assessment method for evaluating structures using this concept. Solutions to such an inverse problem are difficult; data streams from structural measurements may be noisy and sometimes incomplete. Additionally, finding an exact, explicit, closed-form solution to this problem is often impossible. This paper details a neural network approach to solve the inverse impairment detection problem. The approach presented utilizes a deep counter propagation neural network that is capable of modeling input-output functional relations even when mathematically explicit formulas are unavailable or data is noisy and/or corrupt.

A Preprocessed Counterpropagation Neural Network Classifier for Automated Textile Defect Classification

A Preprocessed Counterpropagation Neural Network Classifier for Automated Textile Defect Classification

Automatic musical instrument classification using fractional fourier transform based- MFCC features and counter propagation neural network

This paper presents a novel feature extraction scheme for automatic classification of musical instruments using Fractional Fourier Transform (FrFT)-based Mel Frequency Cepstral Coefficient (MFCC) features. The classifier model for the proposed system has been built using Counter Propagation Neural Network (CPNN). The discriminating capability of the proposed features have been maximized for between-class instruments and minimized for within-class instruments compared to other conventional features. Also, the proposed features show significant improvement in classification accuracy and robustness against Additive White Gaussian Noise (AWGN) compared to other conventional features. McGill University Master Sample (MUMS) sound database has been used to test the performance of the system.

Tool condition monitoring (TCM) using neural networks

Cutting tool conditions significantly influence the quality and precision of the machined parts. With the ability to monitor the cutting tool condition, machining quality can be maintained and catastrophic failure can be eliminated. In this manner, production automation can be achieved. Therefore, tool condition monitoring (TCM) is extremely important to achieve high quality and automation of boring processes. Neural networks have been widely used in condition monitoring. Counterpropagation neural networks (CPNs), which are based on competitive learning, have been utilized in TCM in this research for high quality and automated boring. The inputs of the CPNs were the indexes acquired from three-axis cutting force data. The output was either the tool state or the value of tool wear. Seventy CPN network structures have been utilized for both real-time recognition and real-time measurements. The performance of the CPNs for TCM depends on the network structures. The results of this research are exceedingly successful. Real-time recognition of tool states showed excellent results, using a 2 × 30 × 1 CPN, of being able to predict tool states real time with 100 % accuracy. Real-time measurements can achieve a minimum error of 8.46 % using a 3×69×1 CPN, which is sufficient for continuous assessment of the tool degradation. Control actions can be taken to stop the boring process in order to avoid catastrophic failure and to enhance quality and automation of the boring process.

Three- and four-class classification models for P-glycoprotein inhibitors using counter-propagation neural networks

P-glycoprotein (P-gp) is an ATP binding cassette (ABC) transporter that helps to protect several certain human organs from xenobiotic exposure. This efflux pump is also responsible for multi-drug resistance (MDR), an issue of the chemotherapy approach in the fight against cancer. Therefore, the discovery of P-gp inhibitors is considered one of the most popular strategies to reverse MDR in tumour cells and to improve therapeutic efficacy of commonly used cytotoxic drugs. Until now, several generations of P-gp inhibitors have been developed but they have largely failed in preclinical and clinical studies due to lack of selectivity, poor solubility and severe pharmacokinetic interactions. In this study, three models (SION, SIO, SIN) to classify specific ‘true’ P-gp inhibitors as well as three other models (CPBN, CPB1, CPN) to distinguish between P-gp inhibitors, CYP 3A inhibitors and co-inhibitors of these proteins with rather high accuracy values for the test set and the external set were generated based on counter-propagation neural networks (CPG-NN). Such three and four-class classification models helped provide more information about the bioactivities of compounds not only on one target (P-gp), but also on a combination of multiple targets (P-gp, CYP 3A).

Fuzzy Counter Propagation Neural Network Control for a Class of Nonlinear Dynamical Systems.

Fuzzy Counter Propagation Neural Network (FCPN) controller design is developed, for a class of nonlinear dynamical systems. In this process, the weight connecting between the instar and outstar, that is, input-hidden and hidden-output layer, respectively, is adjusted by using Fuzzy Competitive Learning (FCL). FCL paradigm adopts the principle of learning, which is used to calculate Best Matched Node (BMN) which is proposed. This strategy offers a robust control of nonlinear dynamical systems. FCPN is compared with the existing network like Dynamic Network (DN) and Back Propagation Network (BPN) on the basis of Mean Absolute Error (MAE), Mean Square Error (MSE), Best Fit Rate (BFR), and so forth. It envisages that the proposed FCPN gives better results than DN and BPN. The effectiveness of the proposed FCPN algorithms is demonstrated through simulations of four nonlinear dynamical systems and multiple input and single output (MISO) and a single input and single output (SISO) gas furnace Box-Jenkins time series data.

An Empirical Method for Optimization of Counterpropagation Neural Network Classifier Design for Fabric Defect Inspection

Automated, i.e. machine vision based fabric defect inspection systems have been drawing plenty of attention of the researchers in order to replace manual inspection. Two difficult problems are mainly posed by automated fabric defect inspection systems. They are defect detection and defect classification. Counterpropagation neural network (CPN) is a robust classifier and very promising for defect classification. In general, works reported to date have claimed varying level of successes in detection and classification of different types of defects through CPN; but in particular, no claimed has been made for successful application of CPN for fabric defects detection and classification. In those published works, no investigation has been reported regarding to the variation of major performance parameters of NN based classifiers such as learning time and classification accuracy based on network topology and training parameters. As a result, application engineer has little or no guidance to take design decisions for reaching to optimum structure of NN based defect classifiers in general and CPN based in particular. Our work focuses on empirical investigation of interrelationship between design parameters and performance of CPN based classifier for fabric defect classification. It is believed that such work will be laying the ground to empower application engineers to decide about optimum values of design parameters for realizing most appropriate CPN based classifier.

Intelligent measurements for monitoring and control of glass production furnace for green and efficient manufacturing

Liquefied petroleum (LP) gas is used as one of the fuel systems for glass production furnaces. This research was conducted to develop an intelligent online measurement system for monitoring and control of LP gas so as to achieve green and efficient manufacturing. LP gas is mixed with air at a desired ratio in order to get a proper specific gravity for glass production. Counterpropagation neural networks (CPNs), which are based on competitive learning, were used in this work. Three inputs, air inlet pressure, air/mixed gas differential pressure, and propane/mixed gas differential pressure, were selected for online measurements of specific gravity for monitoring and control of a glass production furnace. Using a 3 × 12 × 1 CPN yields exceedingly successful results for online measurements of specific gravity. An average error of 1.68 %, a minimum error of 0.08 %, and a maximum error of 4.43 % were achieved for online measurements. Control actions can then be taken to achieve much higher energy efficiency which is very important for glass production for green and efficient manufacturing.

Local classification: Locally weighted–partial least squares-discriminant analysis (LW–PLS-DA)

The possibility of devising a simple, flexible and accurate non-linear classification method, by extending the locally weighted partial least squares (LW–PLS) approach to the cases where the algorithm is used in a discriminant way (partial least squares discriminant analysis, PLS-DA), is presented. In particular, to assess which category an unknown sample belongs to, the proposed algorithm operates by identifying which training objects are most similar to the one to be predicted and building a PLS-DA model using these calibration samples only. Moreover, the influence of the selected training samples on the local model can be further modulated by adopting a not uniform distance-based weighting scheme which allows the farthest calibration objects to have less impact than the closest ones.The performances of the proposed locally weighted–partial least squares-discriminant analysis (LW–PLS-DA) algorithm have been tested on three simulated data sets characterized by a varying degree of non-linearity: in all cases, a classification accuracy higher than 99% on external validation samples was achieved. Moreover, when also applied to a real data set (classification of rice varieties), characterized by a high extent of non-linearity, the proposed method provided an average correct classification rate of about 93% on the test set. By the preliminary results, showed in this paper, the performances of the proposed LW–PLS-DA approach have proved to be comparable and in some cases better than those obtained by other non-linear methods (k nearest neighbors, kernel-PLS-DA and, in the case of rice, counterpropagation neural networks).

Application of supervised Kohonen map and counter propagation neural network for classification of nucleic acid structures based on their circular dichroism spectra

One of the most popular instrumental methods to detect the DNA structure is circular dichroism. Specific experimental conditions are required to form different structures of DNA. However, there is the possibility of different structures establishing in the similar circumstance. So, methods development to improve the classification and prediction of structures using their spectra information are needed. To this end, we applied unsupervised (PCA) and supervised (PLS-DA, SKN, and CPNN) approaches to classify CD spectra dataset of different DNA sequences (random coil (ss-DNA), duplex, hairpin, reversed and normal triplex, parallel and antiparallel G-quadruplex, and i-motif). The main part of this work concentrates on the application of artificial neural networks and weight analysis to obtain more classification and prediction accuracy. For this purpose, the trained network was run 10 times, and the average weights were taken. Also, weight analysis was done for the prediction of mixture samples include different structures. The results prove that new method of weights analysis based on SKN and CPNN is useful for classification of complicated data such as different types of DNA structures.

Artificial neural networks to identify naturally existing disease severity status

Classification is a central endeavour in Biology. Heterogeneity of biological systems makes classification more challenging, but this is crucial for effective disease control and management. This study is a computational modelling attempt to classify a plant disease using visual symptoms to ease crop management programmes. Weligama coconut leaf wilt disease (WCLWD), a phytoplasma-borne coconut disease characterised by three foliar symptoms (flaccidity (bending of leaflets), yellowing and marginal necrosis) found in Sri Lanka, was used to demonstrate its applicability. Self-organising map (SOM) was optimised to discover naturally existing categories of WCLWD using foliar symptoms. Ward clustering of SOM identified three distinct disease categories. Results agreed with the nature of disease progression and are supported by K-means clustering. Conversion of SOM clusters to a parsimonious multi-layer perceptron (MLP) supported by a novel efficient network pruning algorithm regenerated identical results proving that precise models can be developed for WCLWD classification using these approaches. The MLP (100 %) outperformed counter propagation (CP) neural network (91 %) in generalisation ability indicating the validity of the MLP model. The study identified flaccidity as the most influential symptom followed by yellowing and necrosis. Comparison of our results with expert decision on disease severity classification revealed 73.45 % correspondence. In-depth investigation into the results from the two approaches using statistical methods revealed that when multiple symptoms are blended, expert decisions rely more on the intensely visible symptom and mainly on a single dimension, whereas the SOM/MLP classifier more accurately captures the average, variation and multi-dimensionality in data indicating that the model is more realistic and capable than the naked eye in detecting the disease.

Self-Organizing Maps for the Classification of Gallic Acylate Polyphenols as HSV-1 Inhibitors

Herpes simplex virus type 1 (HSV-1), a member of the Herpesviridae family, is a ubiquitous, contagious, hostadapted pathogen that causes a wide variety of disease states, such as herpes labialis ("cold sores") and encephalitis. Recently, due to the appearance of acyclovir-resistant HSV-1 mutants, a rapidly growing area of research has been the identification of novel small molecules (whether found in traditional medicine or not) with antiviral activity. One group of these novel pre-drugs is gallic acylate polyphenols. Here, detailed insight into the influence of the chemical structure on anti- HSV-1 activity of gallic acylate polyphenols has been provided based on an exploration of structure-function relationships through self-organizing maps and counterpropagation neural networks. A number of descriptors were investigated to construct optimized models. The resulting model exhibits a correct prediction rate of 90.67%, with active molecule classification accuracy higher than 95.00%, demonstrating that the electrostatic effect and distance between atoms are related to HSV-1 inhibition for these gallic acylate polyphenols. The results provide insights into the influence of the chemical structure on anti-HSV-1 activity of gallic acylate polyphenols.