Evolutionary Synthesis of Logic Circuits Using Information Theory

Aguirre, Arturo Hernández; Coello, Carlos A. Coello

doi:10.1023/b:aire.0000006603.98023.97

Cited by 4 publications

(1 citation statement)

References 32 publications

(29 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the manipulation of directed acyclic graphs, this metaheuristic was the most intuitive for us. It has been used in other domains, for example on graphs representing logic circuits [Aguirre and Coello Coello, 2003]. The design of other metaheuristics in our search space and their comparison with the evolutionary algorithm are left to future work.…”

Section: Search Algorithmmentioning

confidence: 99%

An algorithmic framework for the optimization of deep neural networks architectures and hyperparameters

Keisler¹,

Talbi²,

Claudel³

et al. 2023

Preprint

View full text Add to dashboard Cite

In this paper, we propose an algorithmic framework to automatically generate efficient deep neural networks and optimize their associated hyperparameters. The framework is based on evolving directed acyclic graphs (DAGs), defining a more flexible search space than the existing ones in the literature. It allows mixtures of different classical operations: convolutions, recurrences and dense layers, but also more newfangled operations such as self-attention. Based on this search space we propose neighbourhood and evolution search operators to optimize both the architecture and hyper-parameters of our networks. These search operators can be used with any metaheuristic capable of handling mixed search spaces. We tested our algorithmic framework with an evolutionary algorithm on a time series prediction benchmark. The results demonstrate that our framework was able to find models outperforming the established baseline on numerous datasets.

show abstract

Section: Search Algorithmmentioning

confidence: 99%

An algorithmic framework for the optimization of deep neural networks architectures and hyperparameters

Keisler¹,

Talbi²,

Claudel³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Multi-objective evolutionary design and knowledge discovery of logic circuits based on an adaptive genetic algorithm

Zhao

Jiao

2006

Genet Program Evolvable Mach

View full text Add to dashboard Cite

Evolutionary design of circuits (EDC), an important branch of evolvable hardware which emphasizes circuit design, is a promising way to realize automated design of electronic circuits. In order to improve evolutionary design of logic circuits in efficiency, scalability and capability of optimization, a genetic algorithm based novel approach was developed. It employs a gate-level encoding scheme that allows flexible changes of functions and interconnections of logic cells comprised, and it adopts a multi-objective evaluation mechanism of fitness with weight-vector adaptation and circuit simulation. Besides, it features an adaptation strategy that enables crossover probability and mutation probability to vary with individuals' diversity and genetic-search process. It was validated by the experiments on arithmetic circuits especially digital multipliers, from which a few functionally correct circuits with novel structures, less gate count and higher operating speed were obtained. Some of the evolved circuits are the most efficient or largest ones (in terms of gate count or problem scale) as far as we know. Moreover, some novel and general principles have been discerned from the EDC results, which are easy to verify but difficult to dig out by human experts with existing knowledge. These results argue that the approach is promising and worthy of further research.

show abstract