A clustering approach to incremental learning for feedforward neural networks

Engelbrecht, Andries P.; Brits, R.

doi:10.1109/ijcnn.2001.938474

Cited by 23 publications

(12 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The informativeness is defined as the sensitivity of the neural network output to perturbations in the input value of that pattern [37]. In [36], the output sensitivity vector is defined as (1), S…”

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

“…The incremental learning may also be addressed in context of training data manipulation [35]. For example, in [36], an incremental learning strategy is implemented through the selection of the most informative training samples. Given enough data, an ANN will reconstruct the permittivities of materials and produce a more accurate result.…”

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

“…On the one hand, the characteristic of network should be considered; on the other hand, the sample data must do their best to reflect the intrinsic rule of materials permittivities. The learning strategy is implemented through the selection of most informative training samples [36]. The informativeness is defined as the sensitivity of the neural network output to perturbations in the input value of that pattern [37].…”

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

See 2 more Smart Citations

An Artificial Nerve Network Realization in the Measurement of Material Permittivity

Chen¹,

Huang²,

Yang³

et al. 2011

PIER

View full text Add to dashboard Cite

Abstract-Effective complex permittivity measurements of materials are important in microwave engineering and microwave chemistry. Artificial neural network (ANN) computational module has been used in microwave technology and becomes a useful tool recently. A neural network can be trained to learn the behavior of an effective permittivity of material under microwave irradiation in a test system, and it can provide a fast and accurate result for the permittivity measurement of material. Thus, an on-line measurement has been realized. This paper presents a simple and convenient reconstruction algorithm for determining the dielectric properties of materials. First, a measurement system is designed, and the reflection coefficient is calculated by employing full-wave simulations. Second, an artificial nerve network has been applied, and adequate simulated materials are utilized to train the networks. Last, the trained network is employed to reconstruct the effective permittivity of several organic solvents using the measured scattering parameters, and the reconstructed results for several organic solvents agree well with reference data and the relative errors between them are less than 5%.

show abstract

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

Section: Construction Of the Sample Space [32]mentioning

confidence: 99%

See 1 more Smart Citation

An Artificial Nerve Network Realization in the Measurement of Material Permittivity

Chen¹,

Huang²,

Yang³

et al. 2011

PIER

View full text Add to dashboard Cite

show abstract

“…It avoids relearning of all the parameters by selecting a working subset where the incremental learning is performed. Other incremental learning algorithms include the growing and pruning of classifier architectures [22] and the selection of most informative training samples [23].…”

Section: Related Workmentioning

confidence: 99%

Incremental Multiple Objective Genetic Algorithms

Chen

Guan

2004

IEEE Trans. Syst., Man, Cybern. B

View full text Add to dashboard Cite

Abstract-Incremental learning has been widely addressed in the machine learning literature to cope with learning tasks where the learning environment is ever changing or training samples become available over time. However, most research work explores incremental learning with statistical algorithms or neural networks, rather than evolutionary algorithms. The work in this paper employs genetic algorithms (GAs) as basic learning algorithms for incremental learning within one or more classifier agents in a multiagent environment. Four new approaches with different initialization schemes are proposed. They keep the old solutions and use an "integration" operation to integrate them with new elements to accommodate new attributes, while biased mutation and crossover operations are adopted to further evolve a reinforced solution. The simulation results on benchmark classification data sets show that the proposed approaches can deal with the arrival of new input attributes and integrate them with the original input space. It is also shown that the proposed approaches can be successfully used for incremental learning and improve classification rates as compared to the retraining GA. Possible applications for continuous incremental training and feature selection are also discussed.

show abstract

“…The Incremental Learning may also be addressed in the context of training data manipulation [21]. For instance, in [6], an incremental learning strategy is implemented through the selection of most informative training samples.…”

Section: Introductionmentioning

confidence: 99%

Implementation of Incremental Learning in Artificial Neural Networks

Andrade¹,

Gasca²,

Rendón³

EPiC Series in Computing

View full text Add to dashboard Cite

Nowadays, the use of artificial neural networks (ANN), in particular the Multilayer Perceptron (MLP), is very popular for executing different tasks such as pattern recognition, data mining, and process automation. However, there are still weaknesses in these models when compared with human capabilities. A characteristic of human memory is the ability for learning new concepts without forgetting what we learned in the past, which has been a disadvantage in the field of artificial neural networks. How can we add new knowledge to the network without forgetting what has already been learned, without repeating the exhaustive ANN process? In an exhaustively training is used a complete training set, with all objects of all classes.In this work, we present a novel incremental learning algorithm for the MLP. New knowledge is incorporated into the target network without executing an exhaustive retraining. Objects of a new class integrate this knowledge, which was not included in the training of a source network. The algorithm consists in taking the final weights from the source network, doing a correction of these with the Support Vector Machine tools, and transferring the obtained weights to a target network. This last net is trained with a training set that it is previously preprocessed. The efficiency resulted of the target network is comparable with a net that is exhaustively trained.

show abstract

A clustering approach to incremental learning for feedforward neural networks

Cited by 23 publications

References 11 publications

An Artificial Nerve Network Realization in the Measurement of Material Permittivity

An Artificial Nerve Network Realization in the Measurement of Material Permittivity

Incremental Multiple Objective Genetic Algorithms

Implementation of Incremental Learning in Artificial Neural Networks

Contact Info

Product

Resources

About