Locally Weighted Interpolating Growing Neural Gas

Flentge, Felix

doi:10.1109/tnn.2006.879771

Cited by 11 publications

(7 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4, this maneuver also violates both assumptions of small deviations and decoupled dynamics. Consequently, a classical controller (23) designed specifically for such flight conditions is suboptimal, as can be verified by evaluating the optimality condition (25). For comparison, the neural network controller is updated both with an unconstrained RPROP algorithm [17] and with a constrained RPROP algorithm implementing the adjoined gradient.…”

Section: Constrained Training Implementation and Resultsmentioning

confidence: 99%

“…The ideal control performance over is stated through the optimality condition presented in [40] ( 25) where is the discretized form of (20). This equation constitutes a local condition for optimality that must be satisfied along the system trajectory.…”

Section: ) Online Learning and Short-term Control Knowledgementioning

confidence: 99%

“…2. The critic neural network computes , required by (25), based on the predicted value of the state obtained from (20). Additional design and implementation details are presented in [17].…”

Section: ) Online Learning and Short-term Control Knowledgementioning

confidence: 99%

See 2 more Smart Citations

A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

Ferrari

Jensenius

2008

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

In this paper, a supervised neural network training technique based on constrained optimization is developed for preserving prior knowledge of an input-output mapping during repeated incremental training sessions. The prior knowledge, referred to as long-term memory (LTM), is expressed in the form of equality constraints obtained by means of an algebraic training technique. Incremental training, which may be used to learn new short-term memories (STMs) online, is then formulated as an error minimization problem subject to equality constraints. The solution of this problem is simplified by implementing an adjoined error gradient that circumvents direct substitution and exploits classical backpropagation. A target application is neural network function approximation in adaptive critic designs. For illustrative purposes, constrained training is implemented to update an adaptive critic flight controller, while preserving prior knowledge of an established performance baseline that consists of classical gain-scheduled controllers. It is shown both analytically and numerically that the LTM is accurately preserved while the controller is repeatedly trained over time to assimilate new STMs.

show abstract

Section: Constrained Training Implementation and Resultsmentioning

confidence: 99%

Section: ) Online Learning and Short-term Control Knowledgementioning

confidence: 99%

See 1 more Smart Citation

A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

Ferrari

Jensenius

2008

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

show abstract

“…GNG was originally meant to solve unsupervised learning problems (i.e. clustering and vector quantization); it was extended to supervised RBF networks [53,57] for the incremental generation of neuro-fuzzy systems [59].…”

Section: Model-building With a Modified Growing Neural Gas Networkmentioning

confidence: 99%

MONEDA: scalable multi-objective optimization with a neural network-based estimation of distribution algorithm

et al. 2016

View full text Add to dashboard Cite

The extension of estimation of distribution algorithms (EDAs) to the multiobjective domain has led to multi-objective optimization EDAs (MOEDAs). Most MOEDAs have limited themselves to porting single-objective EDAs to the multi-objective domain. Although MOEDAs have proved to be a valid approach, the last point is an obstacle to the achievement of a significant improvement regarding "standard" multi-objective optimization evolutionary algorithms. Adapting the model-building algorithm is one way to achieve a substantial advance. Most model-building schemes used so far by EDAs employ off-the-shelf machine learning methods. However, the model-building problem has particular requirements that those methods do not meet and even evade. The focus of this paper is on the modelbuilding issue and how it has not been properly understood and addressed by most MOEDAs. We delve down into the roots of this matter and hypothesize about its causes. To gain a deeper understanding of the subject we propose a novel algorithm intended to overcome the drawbacks of current MOEDAs. This new algorithm is the multi-objective neural estimation of distribution algorithm (MONEDA). MONEDA uses a modified growing neural gas network for model-building (MB-GNG). MB-GNG is a custom-made clustering algorithm that meets the above demands. Thanks to its custom-made model-building algorithm, the preservation of elite individuals and its individual replacement scheme, MONEDA is capable of scalably B Luis Martí 123 J Glob Optim solving continuous multi-objective optimization problems. It performs better than similar algorithms in terms of a set of quality indicators and computational resource requirements.

show abstract

“…They have been combined with locally weighted learning methods presented hereafter [36]. The resulting algorithm is endowed with interesting properties for incremental function approximation in a large domain, but to our knowledge it has not been applied to the identification of mechanical models.…”

Section: Self-organizing Mapsmentioning

confidence: 99%

On-line regression algorithms for learning mechanical models of robots: A survey

Sigaud

Salaün

Padois

2011

Robotics and Autonomous Systems

123

View full text Add to dashboard Cite

. On-line regression algorithms for learning mechanical models of robots: a survey. Robotics and Autonomous Systems, Elsevier, 2011, 59 (12) AbstractWith the emergence of more challenging contexts for robotics, the mechanical design of robots is becoming more and more complex. Moreover, their missions often involve unforeseen physical interactions with the environment. To deal with these difficulties, endowing the controllers of the robots with the capability to learn a model of their kinematics and dynamics under changing circumstances is becoming mandatory. This emergent necessity has given rise to a significant amount of research in the Machine Learning community, generating algorithms that address more and more sophisticated on-line modeling questions. In this paper, we provide a survey of the corresponding literature with a focus on the methods rather than on the results. In particular, we provide a unified view of all recent algorithms that outlines their distinctive features and provides a framework for their combination. Finally, we give a prospective account of the evolution of the domain towards more challenging questions.

show abstract

Locally Weighted Interpolating Growing Neural Gas

Cited by 11 publications

References 13 publications

A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

A Constrained Optimization Approach to Preserving Prior Knowledge During Incremental Training

MONEDA: scalable multi-objective optimization with a neural network-based estimation of distribution algorithm

On-line regression algorithms for learning mechanical models of robots: A survey

Contact Info

Product

Resources

About