Adaptive Network Learning Research Articles

A Systematic Kidney Tumour Segmentation and Classification Framework Using Adaptive and Attentive-Based Deep Learning Networks With Improved Crayfish Optimization Algorithm

A Systematic Kidney Tumour Segmentation and Classification Framework Using Adaptive and Attentive-Based Deep Learning Networks With Improved Crayfish Optimization Algorithm

Dynamic adaptive encoder-decoder deep learning networks for multivariate time series forecasting of building energy consumption

Accurate energy consumption prediction models can bring tremendous benefits to building energy efficiency, where the use of data-driven models allows models to be trained based on historical data and to obtain better prediction performance. However, a severe problem has been overlooked where the dynamic spatiotemporal dependence among buildings is often not considered. To settle this issue, this study proposes an asymmetric encoder-decoder learning framework where the spatial relationships and time-series features between multiple buildings are extracted by a convolutional neural network and a gated recurrent neural network to form new input data in the encoder. The decoder then makes predictions based on the input data with an attention mechanism. The model is trained and tested with energy consumption data from 10 different buildings at a university in China. The results show that our proposed model maintains high accuracy using small datasets while predicting different types of buildings, achieving an average R2 of 0.964. Additionally, the accuracy loss in multi-step forecasting is considerably ameliorated by maintaining an average R2 of 0.915 predicting 3 steps ahead of time. Benefiting from the asymmetric encoder-decoder structure, the proposed algorithm resolves the problem of accuracy loss using deep learning in a small dataset and multi-step time series forecasting. Further assisted by the anomaly detection function, the algorithm serves as a reliable guide for the user to effectively manage and control energy consumption in buildings.

Introducing an ecologically-based pest management approach in Cambodia through adaptive learning networks

How can we reach farmers with ecologically-based Integrated Pest Management (IPM) while creating a supportive context for adoption by farmers and relevant stakeholders? We assessed a new method – Adaptive Learning Networks – from reflections of varied stakeholders, farmer diaries and survey data procured in 2016 and 2019. This method is different from current IPM approaches, in that an explicit focus of learning was on engaging other stakeholders to enable adoption by farmers. For example, access to IPM products, providing new services and creating new policies were intrinsically part of the learning process, alongside learning on-farm IPM techniques. The main consideration is to ease farmers from being locked into the practice of pesticide reliance. The method facilitated multi-stakeholder learning that led to an adaptation of the IPM tools and techniques in the case examined. The priority of this new method was socio-technical learning, wherein varied stakeholders modify interactions, incentives and arrangements relating to pest management. Comparing 2016 and 2019 data, there was a significant reduction in insecticide, herbicide, and rodenticide applications. The observed outcomes indicate the potential to enable a wider spread of IPM technologies.

Diffusion Adaptation Strategies for Distributed Estimation Over Gaussian Markov Random Fields

The aim of this paper is to propose diffusion strategies for distributed estimation over adaptive networks, assuming the presence of spatially correlated measurements distributed according to a Gaussian Markov random field (GMRF) model. The proposed methods incorporate prior information about the statistical dependency among observations, while at the same time processing data in real time and in a fully decentralized manner. A detailed mean-square analysis is carried out in order to prove stability and evaluate the steady-state performance of the proposed strategies. Finally, we also illustrate how the proposed techniques can be easily extended in order to incorporate thresholding operators for sparsity recovery applications. Numerical results show the potential advantages of using such techniques for distributed learning in adaptive networks deployed over GMRF.

Adaptive Learning Networks: Developing Resource Management Knowledge through Social Learning Forums

The purpose of this paper is to explore adaptive learning networks as a contemporary means by which new resource management knowledge can develop through social learning forums. The paper draws upon recent discussions within two disparate literatures on indigenous knowledge and network theory and is grounded in fieldwork with two Anishinaabe First Nations in northwestern Ontario. The paper has three objectives. First, problematize the principle of representation as a basic way of including the knowledge of indigenous peoples within natural resource and environmental management. Second, utilize network theory as a way to weave together adaptive learning by individuals into a cross-cultural social learning process. Finally, propose an adaptive natural resources and environmental framework that brings together, through a social learning process, the different ways individuals, indigenous peoples and resource managers, perceive environmental change.

AKL NETWORKS FOR INDUSTRIAL ANALYZER MODELING AND FAULT DETECTION*

An adaptive kernel learning (AKL) networks is proposed to model the industrial analyzer and meanwhile monitor its potential faults, which utilizes kernel function and geometric angle to build an adaptive network topology. Two forms of learning strategies for AKL networks are obtained and their corresponding recursive algorithms are developed, respectively. The proposed AKL algorithm applies to nonlinear MIMO modeling issues with controlled generalization ability. Numerical simulations on Tennessee Eastman (TE) process show that the proposed AKL networks can learn and monitor online the dynamics of the composition analyzer using relatively small samples, under stochastic and fault-existing environment.

A Complex Systems Approach to Learning in Adaptive Networks

A Complex Systems Approach to Learning in Adaptive Networks

Adaptive learning networks of multi-stage fuzzy production rules in expert system of grinding characteristics

Abstract The cross validation technique offers a criterion to measure the degree of approximation of a mathematical model. The advantage of the leaving-one-out error estimation and the k-fold cross validation is that almost all the available samples are used for training whereas all the samples are used for testing. But because these techniques are computationally expensive, it has often been reserved for problems with small sample size. This paper discusses the validity of Akaike's AIC for selecting the number of layers in an adaptive learning network of GMDH type whose partial descriptions are represented by Gaussian functions. In numerical examples, several computer simulations of learning and comparisons of AIC with cross validation procedure are shown. The expert system for identification of grinding characteristics is discussed.

Adaptive learning networks in APL2

The paper considers Adaptive Learning Networks (ALN) as a tool to solve the problems of modeling, prediction, diagnostics and pattern recognition in complex systems. This method is similar to the neural network technique. The main difference is the self-organization of network structure on the basis of generation and estimation of various nodes, connections and weights. A set of functions presented in the paper shows that ALNs are easily realized in APL2. User-defined operators are used as a very convenient tool for ALN programming. The paper discusses the application of implemented software to the problem of Burnout Heat Flux Prediction in nuclear reactors. It is shown that ALN technique allows the prediction of burnout heat flux with approximately three times better accuracy than other commonly used methods.

Cue Competition Effects: Empirical Tests of Adaptive Network Learning Models

The ability to predict future consequences on the basis of previous experience with the current set of environmental cues is one of the most fundamental of all cognitive processes. This study investigated how the validity of one cue influences the effectiveness of another cue for predicting a criterion. The results demonstrate a cue competition effect—increasing the validity of one cue decreased the effectiveness of another cue in a linear prediction task, even though the two cues were statistically independent.

Cue Competition Effects: Theoretical Implications for Adaptive Network Learning Models

A feature-free method for testing adaptive network learning models is presented. The test is based on a property called the mean matching law, a. property shared by many adaptive network models of learning. As an application of this method, we prove that cue competition effects obtained with statistically independent cues cannot be explained by many previous adaptive network learning models, including those based on the delta learning rule. These results point to the need to incorporate competitive learning properties into adaptive network learning models.

Adaptive networks for fault diagnosis and process control

The use of adaptive (artificial neural) networks for fault diagnosis and process control is explored. Adaptive networks can be used as fault recognition systems, as adaptive nonlinear process models, and as controllers. Connection strengths representing correlations between inputs (alarms and sensor measurements) and outputs (faults, future sensor measurements or control actions) are learned using the LMS (Widrow-Hoff) rule and the backpropagation algorithm. The resulting system is a pattern recognizer which is able to learn nonlinear and logical relationships as well as linear correlations. Results are presented for two problems: diagnosing failures in a small model chemical plant, and controlling a highly nonlinear bioreactor. For the diagnosis problem, learning in adaptive networks given qualitative (alarm) and quantitative (sensor) data are compared, and the effect of noise is studied. The importance of nonlinear networks is demonstrated using simple problems which require context sensitivity and problems where optimal alarm thresholds are learned. Results using an adaptive model-based controller using two neural networks (one for the model and one for the controller) are presented and extensions to the standard layered feedforward network are suggested which greatly increase the utility of neural networks for process control. Adaptive networks provide pattern recognition facilities which can be used and interpreted in several ways: they perform multiple nonlinear regression on input/output pairs (current sensor readings and alarms are associated with faults, future readings or control actions) that may be both quantitative and qualitative. Although adaptive networks have many similarities with well-established statistical techniques for system identification, they still offer promise of major benefit, primarily in suggesting new equations, architectures and algorithms. Although the adaptive networks can be viewed as a special form of nonlinear regression, the network formalism suggests several powerful nonlinear functional forms to use. Use of highly interconnected nonlinear systems allows unexpected interactions to be captured. Recurrent networks can learn to recognize arbitrary delays. Temporal difference methods can speed learning when feedback is not immediate. Artificial neural networks will not in any way replace control algorithms, but rather are good for learning that which we are ignorant of: alarm thresholds, patterns of disturbances and model-mismatch (including process delays). Adaptive networks can be thought of as a very data-intensive approach to system identification: many parameters are used in a format that allows interactions between all of the variables. Thus more specific patterns can be learned than when the system is described using a relatively simple equation with a small number of parameters. When the exact form of the equation is known or little data is available, an equation with fewer parameters is of course preferable and neural networks should be avoided. We looked at two example problems: fault diagnosis and adaptive model-based control. Both quantitative (sensor) and qualitative (alarm) information can be used in fault diagnosis. Optimal thresholds for triggering alarms are learned; These can be dependent on the context provided by the states of other variables. Nonlinear networks are required for all but the simplest problems. In adaptive model-based control, a highly nonlinear model was learned without using a priori knowledge of the equational forms. This approach is expected to yield the most benefit in MIMO systems which contain complex nonlinear interactions and in systems in which recurring disturbances must be recognized and forecast.

Hebbian Synapses: Biophysical Mechanisms and Algorithms

We have examined the evolution of the concept of a Hebbian synaptic modification and have suggested a contemporary definition. The biophysical mechanism demonstrated in vitro to control the induction of one type of hippocampal LTP has been shown to satisfy our definition of a Hebbian synaptic modification. Whether this biophysical mechanism is involved in the organization of behavior in the manner that Hebb originally envisioned remains to be seen. We have also summarized several modification algorithms that have been explored in theoretical studies of learning in adaptive networks. These algorithms also satisfied our definition of a Hebbian modification, but their relationships to known neurobiology require further exploration. By reviewing the biophysical mechanisms and formal algorithms together, we have exposed obvious similarities and differences. Such comparisons may help bridge the gap between computational theory and knowledge of the neurobiology of use-dependent synaptic change. Current models of LTP reveal that the activity-modification relationships are extremely sensitive to the biophysical/molecular details. The activity-modification relationships obviously can have a major influence on adaptive neurodynamics at the network level. As more accurate representations of the biological complexity and diversity are introduced into adaptive network simulations, we expect to gain new insights into the classes of computation that particular networks are capable of performing.

Application of adaptive learning networks for the detection of failing power system components

A system capable of monitoring mine electrical power systems and detecting component failure in early stages could significantly improve power system safety and availability. Such monitoring would require a method of evaluating electrical features, calculated from terminal values, for indications of component deterioration. Research is being conducted by the US Bureau of Mines to examine the use of mathematical models to aid in this evaluation by creating polynomial networks called adaptive learning networks that can indicate deteriorated conditions in cable-connected motor systems. This process uses laboratory training data to select the electrical features most significant for accurately modeling cable-motor system conditions and forms mathematical expressions relating these features to the presence and severity of deterioration. Particular attention is given to PNETTR-4X, a modelling software package that forms adaptive learning networks from input training data using an unsupervised learning process. Models developed thus far can process readily measured terminal information and quantify deterioration power and current to within 3% of motor full load values. >

Feature selection using adaptive learning networks for text-independent speaker verification

A nonlinear transformation which makes use of the polynomial discriminant function (PDF) is applied to the selection of features for text-independent speaker verification. For each speaker within the set, a PDF is established such that it maps the original attribute set onto the one-dimensional one resulting in the best discrimination of the speaker from the others in the new feature space. The complicated PDF is implemented using the technique known as the adaptive learning network (ALN). In this case, the multinomial discriminant function is estimated using interconnecting fundamental building blocks each of which computes a two-element quadratic discriminant function. Through a comprehensive training procedure, the coefficients needed to describe each building block and the interconnecting configurations of them are determined. Application of the chosen attributes to a text-independent speaker verification experiment yields relatively low false speaker rejection and verification error rates.

Evaluation and comparison of pattern classifiers for chemical applications: adaptive digital learning networks and linear discriminants

ADVERTISEMENT RETURN TO ISSUEPREVArticleNEXTEvaluation and comparison of pattern classifiers for chemical applications: adaptive digital learning networks and linear discriminantsLeonard J. Soltzberg, Charles L. Wilkins, Steven L. Kaberline, T. Fai Lam, and T. L. BrunnerCite this: J. Am. Chem. Soc. 1976, 98, 23, 7144–7151Publication Date (Print):October 1, 1976Publication History Published online1 May 2002Published inissue 1 October 1976https://doi.org/10.1021/ja00439a005RIGHTS & PERMISSIONSArticle Views42Altmetric-Citations6LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InReddit PDF (833 KB) Get e-Alerts Get e-Alerts

Classification of mass spectra using adaptive digital learning networks

Classification of mass spectra using adaptive digital learning networks