Convolutional Neural Networks (CNNs) have gained a remarkable success on many image classification tasks in recent years. However, the performance of CNNs highly relies upon their architectures. For most state-of-the-art CNNs, architectures are often manually-designed with expertise both and investigated problems. Therefore, it is difficult for users, who no extended to design optimal CNN own problems interest. In this paper, we propose an automatic architecture method by using genetic...

Evolutionary paradigms have been successfully applied to neural network designs for two decades. Unfortunately, these methods cannot scale well the modern deep networks due complicated architectures and large quantities of connection weights. In this paper, we propose a new method using genetic algorithms evolving weight initialization values convolutional address image classification problems. proposed algorithm, an efficient variable-length gene encoding strategy is designed represent...

Condition monitoring of dynamic systems based on vibration signatures has generally relied upon Fourier-based analysis as a means translating signals in the time domain into frequency domain. However, Fourier provided poor representation well localized time. In this case, it is difficult to detect and identify signal pattern from expansion coefficients because information diluted across whole basis. The wavelet packet transform (WPT) introduced an alternative extracting time-frequency...

Inverted generational distance (IGD) has been widely considered as a reliable performance indicator to concurrently quantify the convergence and diversity of multiobjective many-objective evolutionary algorithms. In this paper, an IGD indicator-based algorithm for solving optimization problems (MaOPs) proposed. Specifically, is employed in each generation select solutions with favorable diversity. addition, computationally efficient dominance comparison method designed assign rank values...

Deep Neural Networks (DNNs) have achieved great success in many applications. The architectures of DNNs play a crucial role their performance, which is usually manually designed with rich expertise. However, such design process labour intensive because the trial-and-error process, and also not easy to realize due rare expertise practice. Architecture Search (NAS) type technology that can automatically. Among different methods NAS, Evolutionary Computation (EC) recently gained much attention...

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> This paper proposes a constraint handling technique for multiobjective evolutionary algorithms based on an adaptive penalty function and distance measure. These two functions vary dependent upon the objective value sum of violations individual. Through this design, space is modified to account performance violation each The are used in nondominance sorting facilitate search optimal solutions not...

Evolutionary algorithms have been effectively used to solve multiobjective optimization problems with a small number of objectives, two or three in general. However, when many objectives are encountered, nearly all perform poorly due loss selection pressure fitness evaluation solely based upon the Pareto optimality principle. In this paper, we introduce new mechanism continuously differentiate individuals into different degrees beyond classification original dominance. The concept fuzzy...

The performance of convolutional neural networks (CNNs) highly relies on their architectures. In order to design a CNN with promising performance, extensive expertise in both CNNs and the investigated problem domain is required, which not necessarily available every interested user. To address this problem, we propose automatically evolve architectures by using genetic algorithm (GA) based ResNet DenseNet blocks. proposed completely automatic designing particular, neither preprocessing...

Managing convergence and diversity is essential in the design of multiobjective particle swarm optimization (MOPSO) search an accurate well distributed approximation true Pareto-optimal front. Largely due to its fast convergence, incurs a rapid loss during evolutionary process. Many mechanisms have been proposed existing MOPSOs terms leader selection, archive maintenance, perturbation tackle this deficiency. However, few are designed dynamically adjust balance exploration exploitation...

There have been few researches on solving multimodal multiobjective optimization problems, whereas they are commonly seen in real-world applications but difficult for the existing evolutionary optimizers. In this paper, we propose a novel algorithm using two-archive and recombination strategies. proposed algorithm, properties of decision variables relationships among them analyzed at first to guide search. Then, general framework two archives, i.e., convergence diversity is adopted...

Convolutional neural networks (CNNs) have shown remarkable performance in various real-world applications. Unfortunately, the promising of CNNs can be achieved only when their architectures are optimally constructed. The state-of-the-art typically handcrafted with extensive expertise both and investigated data, which consequently hampers widespread adoption for less experienced users. Evolutionary deep learning (EDL) is able to automatically design best CNN without much expertise. However,...

When solving constrained optimization problems by evolutionary algorithms, an important issue is how to balance constraints and objective function. This paper presents a new method address the above issue. In our method, after generating offspring for each parent in population making use of differential evolution (DE), well-known feasibility rule used compare its parent. Since prefers function, function information has been exploited as follows: if cannot survive into next generation value...

Deep Learning (DL) aims at learning the \emph{meaningful representations}. A meaningful representation refers to one that gives rise significant performance improvement of associated Machine (ML) tasks by replacing raw data as input. However, optimal architecture design and model parameter estimation in DL algorithms are widely considered be intractable. Evolutionary much preferable for complex non-convex problems due its inherent characteristics gradient-free insensitivity local optimum. In...

Dynamic multiobjective optimization problems (DMOPs) are with multiple conflicting objectives, and these objectives change over time. Transfer learning-based approaches have been proven to be promising; however, a slow solving speed is one of the main obstacles preventing such methods from real-world problems. One reasons for running that low-quality individuals occupy large amount computing resources, may lead negative transfer. Combining high-quality individuals, as knee points, transfer...

In this paper, we propose a generic, two-phase framework for solving constrained optimization problems using genetic algorithms. the first phase of algorithm, objective function is completely disregarded and problem treated as constraint satisfaction problem. The search directed toward minimizing violation solutions eventually finding feasible solution. A linear rank-based approach used to assign fitness values individuals. solution with least archived elite in population. second phase,...

This paper proposes a self adaptive penalty function for solving constrained optimization problems using genetic algorithms. In the proposed method, new fitness value, called distance in normalized fitness-constraint violation space, and two values are applied to infeasible individuals so that algorithm would be able identify best current population. The method aims encourage with low objective value constraint violation. number of feasible population is used guide search process either...

Abstract Research fronts, defined as clusters of documents that tend to cite a fixed, time invariant set base documents, are plotted lines for visualization and exploration. Using related the subject anthrax research, this article illustrates construction, exploration, interpretation purpose identifying visualizing temporal changes in research activity through journal articles. Such information is useful presentation members expert panels used technology forecasting.

This paper proposes an adaptive penalty function for solving constrained optimization problems using genetic algorithms. The proposed method aims to exploit infeasible individuals with low objective value and constraint violation. number of feasible in the population is used guide search process either toward finding more or searching optimum solution. simple implement does not need any parameter tuning. performance algorithm tested on 22 benchmark functions literature. results show that...

This paper proposes a new evolutionary approach to multiobjective optimization problems - the dynamic algorithm (DMOEA). In DMOEA, novel cell-based rank and density estimation strategy is proposed efficiently compute dominance diversity information when population size varies dynamically. addition, growing declining strategies are designed determine if an individual will survive or be eliminated based on some qualitative indicators. Meanwhile, objective space compression devised continuously...

Multimodal optimization is one of the most challenging tasks for optimization. It requires an algorithm to effectively locate multiple global and local optima, not just single optimum as in a objective problem. To address this objective, paper first investigates cluster-based differential evolution (DE) multimodal problems. The clustering partition used divide whole population into subpopulations so that different can optima. Furthermore, self-adaptive parameter control employed enhance...

Evolutionary algorithms have been successfully exploited to solve multiobjective optimization problems. In the literature, a heuristic approach is often taken. For chosen benchmark problem with specific characteristics, performance of evolutionary (MOEAs) evaluated via some metrics. The conclusion then drawn based on statistical findings given preferable choices conclusion, if any, indecisive and reveals no insight pertaining which characteristics underlying MOEA could perform best. this...

In the field of evolutionary computation, there has been a growing interest in applying algorithms to solve multimodal optimization problems (MMOPs). Due fact that an MMOP involves multiple optimal solutions, many niching methods have suggested and incorporated into for locating such solutions single run. this paper, we propose novel transformation technique based on multiobjective MMOPs, called MOMMOP. MOMMOP transforms problem with two conflicting objectives. After above transformation,...

Recently, various multiobjective particle swarm optimization (MOPSO) algorithms have been developed to efficiently and effectively solve problems. However, the existing MOPSO designs generally adopt a notion "estimate" fixed population size sufficiently explore search space without incurring excessive computational complexity. To address issue, this paper proposes integration of dynamic strategy within multiple-swarm MOPSO. The proposed algorithm is named An additional feature, adaptive...