arXiv subjects

Yuichi Nagata

Publications and source records attributed to Yuichi Nagata.

5 recordsLinked to original sources

arXiv2026

Associative Constructive Evolution: Enhancing Metaheuristics through Hebbian-Learned Generative Guidance

Metaheuristic algorithms such as Particle Swarm Optimization (PSO) and Evolutionary Algorithms (EA) excel at exploring solution spaces but lack mechanisms to accumulate and reuse procedural knowledge from successful search trajectories. This paper proposes Associative Constructive Evolution (ACE), a framework that enhances metaheuristics through learned generative guidance. ACE introduces a Generative Construction Automaton (GCA) -- a probabilistic model over operation sequences -- coupled with the base metaheuristic in a synergistic loop: the metaheuristic explores and provides trajectory samples, while the GCA consolidates successful patterns and guides future exploration. Three mechanisms realize this cooperation: Hebbian weight consolidation that strengthens associations between co-successful operations, guided sampling that biases search toward learned high-quality regions, and symbolic abstraction that extracts frequent patterns into reusable macro-operations. Experiments integrating ACE with EA and PSO on molecular design and maze navigation demonstrate consistent improvements. ACE-PSO achieves a 27.5% increase in success rate while reducing convergence time by 49.6%. In molecular design, ACE-EA improves fitness by 10.1% with 126 chemically interpretable macro-operations automatically discovered.

cs.NE↗

arXiv2026

Multi-Constrained Evolutionary Molecular Design Framework: An Interpretable Drug Design Method Combining Rule-Based Evolution and Molecular Crossover

This study proposes MCEMOL (Multi-Constrained Evolutionary Molecular Design Framework), a molecular optimization approach integrating rule-based evolution with molecular crossover. MCEMOL employs dual-layer evolution: optimizing transformation rules at rule level while applying crossover and mutation to molecular structures. Unlike deep learning methods requiring large datasets and extensive training, our algorithm evolves efficiently from minimal starting molecules with low computational overhead. The framework incorporates message-passing neural networks and comprehensive chemical constraints, ensuring efficient and interpretable molecular design. Experimental results demonstrate that MCEMOL provides transparent design pathways through its evolutionary mechanism while generating valid, diverse, target-compliant molecules. The framework achieves 100% molecular validity with high structural diversity and excellent drug-likeness compliance, showing strong performance in symmetry constraints, pharmacophore optimization, and stereochemical integrity. Unlike black-box methods, MCEMOL delivers dual value: interpretable transformation rules researchers can understand and trust, alongside high-quality molecular libraries for practical applications. This establishes a paradigm where interpretable AI-driven drug design and effective molecular generation are achieved simultaneously, bridging the gap between computational innovation and practical drug discovery needs.

cs.NE↗

arXiv2020

Escherization with Generalized Distance Functions Focusing on Local Structural Similarity

The Escherization problem involves finding a closed figure that tiles the plane that is most similar to a given goal figure. In Koizumi and Sugihara's formulation of the Escherization problem, the tile and goal figures are represented as $n$-point polygons where the similarity between them is measured based on the difference in the positions between the corresponding points. This paper presents alternative similarity measures (distance functions) suitable for this problem. The proposed distance functions focus on the similarity of local structures in several different manners. The designed distance functions are incorporated into a recently developed framework of the exhaustive search of the templates for the Escherization problem. Efficient exhaustive and incomplete search algorithms for the formulated problems are also developed to obtain results within a reasonable computation time. Experimental results showed that the proposed algorithms found satisfactory tile shapes for fairly complicated goal figures in a reasonable computation time.

cs.CG↗

arXiv2019

An Efficient Algorithm for the Escherization Problem in the Polygon Representation

In the Escherization problem, given a closed figure in a plane, the objective is to find a closed figure that is as close as possible to the input figure and tiles the plane. Koizumi and Sugihara's formulation reduces this problem to an eigenvalue problem in which the tile and input figures are represented as $n$-point polygons. In their formulation, the same number of points are assigned to every tiling edge, which forms a tiling template, to parameterize the tile shape. By considering all possible configurations for the assignment of the $n$ points to the tiling edges, we can achieve much flexibility in terms of the possible tile shapes and the quality of the optimal tile shape improves drastically, at the cost of enormous computational effort. In this paper, we propose an efficient algorithm to find the optimal tile shape for this extended formulation of the Escherization problem.

cs.CG↗

arXiv2012

Theoretical foundation for CMA-ES from information geometric perspective

This paper explores the theoretical basis of the covariance matrix adaptation evolution strategy (CMA-ES) from the information geometry viewpoint. To establish a theoretical foundation for the CMA-ES, we focus on a geometric structure of a Riemannian manifold of probability distributions equipped with the Fisher metric. We define a function on the manifold which is the expectation of fitness over the sampling distribution, and regard the goal of update of the parameters of sampling distribution in the CMA-ES as maximization of the expected fitness. We investigate the steepest ascent learning for the expected fitness maximization, where the steepest ascent direction is given by the natural gradient, which is the product of the inverse of the Fisher information matrix and the conventional gradient of the function. Our first result is that we can obtain under some types of parameterization of multivariate normal distribution the natural gradient of the expected fitness without the need for inversion of the Fisher information matrix. We find that the update of the distribution parameters in the CMA-ES is the same as natural gradient learning for expected fitness maximization. Our second result is that we derive the range of learning rates such that a step in the direction of the exact natural gradient improves the parameters in the expected fitness. We see from the close relation between the CMA-ES and natural gradient learning that the default setting of learning rates in the CMA-ES seems suitable in terms of monotone improvement in expected fitness. Then, we discuss the relation to the expectation-maximization framework and provide an information geometric interpretation of the CMA-ES.

cs.NE↗