Masayuki Yamamura

A genetic algorithm for job-shop scheduling problems using job-based order crossover

We propose a genetic algorithm for job shop scheduling problems. The proposed method uses a job sequence matrix. This paper introduces a new crossover, the job based order crossover (JOX), which can preserve characteristics very well. JOX preserves the order of each job on all machines between parents and their children, taking account of the dependency among machines. Since the children generated by JOX are not always feasible, we propose a technique to transform them into active schedules by using the Giffler and Thompson method (B. Giffler and G.L. Thompson, 1969). Furthermore, we introduce a mutation for maintaining a diversity of population without disrupting characteristics. By applying the proposed method to Fisher and Thompsons 10/spl times/10 and 20/spl times/5 problems (H. Fisher and G.L. Thompson, 1963), we show its usefulness.

Theoretical Analysis of Simplex Crossover for Real-Coded Genetic Algorithms

In this paper, we perform theoretical analysis and experiments on the Simplex Crossover (SPX), which we have proposed. Real-coded GAs are expected to be a powerful function optimization technique for real-world applications where it is often hard to formulate the objective function. However, we believe there axe two problems which will make such applications difficult; 1) performance of real-coded GAs depends on the coordinate system used to express the objective function, and 2) it costs much labor to adjust parameters so that the GAs always find an optimum point efficiently. The result of our theoretical analysis and experiments shows that a performance of SPX is independent of linear coordinate transformation and that SPX always optimizes various test function efficiently when theoretical value for expansion rate, which is a parameter of SPX, is applied.

Aqueous computing: writing on molecules

Molecular computing is viewed here as a process of writing on molecules while they are dissolved in water. When DNA molecules are employed, they are used only in double stranded form and only as data registers. All computations are initialized with the same single molecular variety. Current progress toward laboratory prototyping of computations is reported.

Self-organizing formation algorithm for active elements

We propose a novel method of self-organizing formation. It is assumed that elements are not connected to each other and they can move in continuous space. The objective is to arrange elements in a certain spatial pattern like a crystal, and to make the outline of the group in the desired shape. For this purpose, we propose a method by using virtual springs among the elements. In this algorithm, an element generates virtual springs between the neighbor element based on information of how many other elements exist in the neighborhood with a certain radius. Although the elements interact locally, only by virtual springs, and they do not have global information at all, they form a shape much larger than the sensory radius. By a simulation study, we confirmed convergence to a target shape from a random state in very high probability. This kind of algorithm gives a new principle of self-organizing formation, and its simplicity will be useful for the design of self-assembling nano machines in future.

Multitask reinforcement learning on the distribution of MDPs

In this paper we address a new problem in reinforcement learning. Here we consider an agent that faces multiple learning tasks within its lifetime. The agents objective is to maximize its total reward in the lifetime as well as a conventional return in each task. To realize this, it has to be endowed an important ability to keep its past learning experiences and utilize them for improving future learning performance. This time we try to phrase this problem formally. The central idea is to introduce an environmental class, BV-MDPs that is defined with the distribution of MDPs. As an approach to exploiting past learning experiences, we focus on statistics (mean and deviation) about the agents value tables. The mean can be used as initial values of the table when a new task is presented. The deviation can be viewed as measuring reliability of the mean, and we utilize it in calculating priority of simulated backups. We conduct experiments in computer simulation to evaluate the effectiveness.

k -Certainty exploration method: an action selector to identify the environment in reinforcement learning

Copyright (c) 1996 Elsevier Science B.V. All rights reserved. Reinforcement learning aims to adapt an agent to an unknown environment according to rewards. There are two issues to handle delayed reward and uncertainty. Q-learning is a representative reinforcement learning method. It is used in many works since it can learn an optimum policy. However, Q-learning needs numerous trials to converge to an optimum policy. If the target environments can be described in Markov decision processes, we can identify them from statistics of sensor−action pairs. When we build the correct environment model, we can derive an optimum policy with the Policy Iteration Algorithm. Therefore, we can construct an optimum policy through identifying environments efficiently. We separate the learning process into two phases: identifying an environment and determining an optimum policy. We propose the k-Certainty Exploration Method for identifying an environment. After that, an optimum policy is determined by the Policy Iteration Algorithm. We call a rule k-certainty if and only if it has been selected k times or more. The k-Certainty Exploration Method excepts any loop of rules that already achieve k-certainty. We show its effectiveness by comparing it with Q-learning in two experiments. One is Suttons maze-like environment, the other is an original environment where an optimum policy varies according to a parameter.

Tunable synthetic phenotypic diversification on Waddington’s landscape through autonomous signaling

Phenotypic diversification of cells is crucial for developmental and regenerative processes in multicellular organisms. The diversification concept is described as the motion of marbles rolling down Waddington’s landscape, in which the number of stable states changes as development proceeds. In contrast to this simple concept, the complexity of natural biomolecular processes prevents comprehension of their design principles. We have constructed, in Escherichia coli, a synthetic circuit with just four genes, which programs cells to autonomously diversify as the motion on the landscape through cell–cell communication. The circuit design was based on the combination of a bistable toggle switch with an intercellular signaling system. The cells with the circuit diversified into two distinct cell states, “high” and “low,” in vivo and in silico, when all of the cells started from the low state. The synthetic diversification was affected by not only the shape of the landscape determined by the circuit design, which includes the synthesis rate of the signaling molecule, but also the number of cells in the experiments. This cell-number dependency is reminiscent of the “community effect”: The fates of developing cells are determined by their number. Our synthetic circuit could be a model system for studying diversification and differentiation in higher organisms. Prospectively, further integrations of our circuit with different cellular functions will provide unique tools for directing cell fates on the population level in tissue engineering.

A robust real-coded evolutionary algorithm with toroidal search space conversion

This paper presents a robust Real-coded evolutionary algorithm. Real-coded evolutionary algorithms (RCEAs), such as real-coded genetic algorithms and evolution strategies, are known as effective methods for function optimization. However, they often fail to find the optimum in case the objective function is multimodal, discrete or high-dimensional. It is also reported that most crossover (or recombination) operators for RCEAs has sampling bias that prevents to find the optimum near the boundary of search space. They like to search the center of search space much more than the other. Therefore, they will not work on functions that have their optima near the boundary of search space. In this paper, we apply two methods, genetic algorithm with search area adaptation (GSA) and toroidal search space conversion (TSC), to the function optimization for improving the robustness of RCEAs. The former method searches adaptively and the latter one removes the sampling bias. Through several experiments, we have confirmed that GSA works adaptively and it shows higher performance, and RCEAs with TSC show effectiveness to find the optima near the boundary of search space.

Aqueous computing: a survey with an invitation to participate

The concept of aqueous computing is presented here, first in full generality, and afterward, using an implementation in a specific enzymatic technology. Aqueous computing arose in the context of biomolecular (DNA) computing, but the concept is independent of the specifics of its biochemical origin. Alternate technologies for realizing aqueous computing are being considered for future implementation. A solution of an instance of the Boolean satisfiability problem, (SAT), is reported here that provides a new example of an aqueous computation that has been carried out successfully. This small instance of the SAT problem is sufficiently complex to allow our current enzymatic technology to be illustrated in detail. The reader is invited to participate in the rich interdisciplinary activity required by wet lab computing. A project is suggested to the reader for determining the three-colorings of a graph. The basic operations required for this project are exhibited in the solution of the SAT example reported here.

Aqueous Solutions of Algorithmic Problems: Emphasizing Knights on a 3 x 3

A pattern for performing several DNA computations is outlined using the aqueous approach, the essence of which is writing on molecules dissolved in water. Four of the indicated computations have been carried out in wet labs in the aqueous style. As an illustration, gel photos will be exhibited that confirm the correctness of a small SAT computation. Emphasis will be placed on the aqueous approach, now in progress, to the problem of producing the set of all patterns in which knights can be placed on a 3 � 3 chessboard with no knight attacking another. Currently the writing technology used is based on molecular biology. In the future we hope that light can replace biochemistry as the writing procedure.