2009
DOI: 10.1080/10426910802678321
|View full text |Cite
|
Sign up to set email alerts
|

Identification and Optimization of AB2Phases Using Principal Component Analysis, Evolutionary Neural Nets, and Multiobjective Genetic Algorithms

Abstract: Available data for a large number of AB 2 compounds were subjected to a rigorous study using a combination of Principal Component Analysis (PCA) technique, multiobjective genetic algorithms, and neural networks that evolved through genetic algorithms. The identification of various phases and phase-groups were very successfully done using a decision tree approach. Since the variable hyperspaces for the different phases were highly intersecting in nature, a cumulative probability index was defined for the format… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
10
0

Year Published

2010
2010
2024
2024

Publication Types

Select...
5
2

Relationship

2
5

Authors

Journals

citations
Cited by 26 publications
(10 citation statements)
references
References 29 publications
0
10
0
Order By: Relevance
“…In this study the outputs of the FLUENT™ calculations were fed to an evolutionary neural network 16 that itself evolved as a bi‐objective tradeoff between the training error and the complexity of network, measured by the total number of connections in the lower part of the network, including the biases. The upper part evolved through a Linear Least Square technique and a Predator Prey Genetic Algorithm was employed for the bi‐objective optimization task, as in many of our earlier work 16, 18–20. In this procedure we dealt with a population of neural nets, of which the random parents were subjected to a crossover operation to produce the children networks .…”
Section: Modeling the Castermentioning
confidence: 99%
“…In this study the outputs of the FLUENT™ calculations were fed to an evolutionary neural network 16 that itself evolved as a bi‐objective tradeoff between the training error and the complexity of network, measured by the total number of connections in the lower part of the network, including the biases. The upper part evolved through a Linear Least Square technique and a Predator Prey Genetic Algorithm was employed for the bi‐objective optimization task, as in many of our earlier work 16, 18–20. In this procedure we dealt with a population of neural nets, of which the random parents were subjected to a crossover operation to produce the children networks .…”
Section: Modeling the Castermentioning
confidence: 99%
“…A good example of this was the recent identification and optimization of inorganic AB 2 -Laves phases for H-storage (Agarwal et al, 2009;Rajagopalan et al, 2003). Such new hybrid approaches link data dimensionality reduction methods to evolutionary algorithms to uncover correlations masked by solely using GAs.…”
Section: Further Readingmentioning
confidence: 98%
“…Such new hybrid approaches link data dimensionality reduction methods to evolutionary algorithms to uncover correlations masked by solely using GAs. Coupling PCA, predatorÀprey GA, and neural nets, compounds were classified effectively according to distorted packing sequences, with data mining never having discovered such simple classifications (Agarwal et al, 2009). For structure predictions, evolutionary algorithms have also been utilized in various forms.…”
Section: Further Readingmentioning
confidence: 99%
“…So now we can add another level of complexity into our analysis by considering how the strength of their influence may actually change as we add more data. Agarwal et al (48) explored this issue by developing a hybrid algorithmic approach that carefully linked a variety of techniques, including evolutionary neural nets and multiobjective genetic algorithms. Of particular interest is the embedding of predator-prey algorithms with genetic algorithmic methods, whereby these researchers explored how the descriptors (genes) can evolve and interact dynamically.…”
Section: Looking For a Robust Genementioning
confidence: 99%