Search

Aiming at the problems of low recognition accuracy and large calculation amount of age estimation, this paper proposes a staged single aggregation backbone network VoVNetv4. In order to improve the feature extraction ability of the network, the Regional One-Shot Aggregation (ROSA) module is proposed in the network for regional encapsulation of the feature module. The classifier adopts the COnsistent RAnk Logits (CORAL) framework, which can solve the problem of inconsistency of classifiers in ordered regression CNNs for multiple binary classification tasks. This paper conducts experimental verification on the Asian Face Age Dataset (AFAD) and compares it with other research methods. The experimental results show that VoVNetv4 can effectively improve the feature extraction and detection ability of the network and improve the detection efficiency of age estimation. [ABSTRACT FROM AUTHOR]

Machine learning (ML) is evolving rapidly and has made many theoretical breakthroughs while widely applied in various fields. ML allows systems the ability to access data and use it to enable computers to execute cognitive processes such as learning and improving from previous experiences and solving complicated issues. Many first-order stochastic optimization methods have been used to solve the optimization model of ML. These algorithms adopt Barzilai-Borwein (BB) step size instead of fixed or diminishing step size to improve performance. However, the BB step size format involves fractional calculation, which inevitably leads to a zero denominator, especially when the objective function is non-convex. The BB technique will be violated if the denominator is near 0 or even negative. To improve the computation of the step size, a Positive Defined Stabilized Barzilai-Borwein (PDSBB) approach is introduced in this paper. Integrating PDSBB with the stochastic variance reduced gradient (SVRG) approach, a new method SVRG-PDSBB is proposed. Numerical experiments have shown that the new algorithm has stabilized the performance of the new step size, which successfully avoiding zero denominators and effectively solving the common problems in machine learning. The convergence of SVRG-PDSBB is theoretically and numerically proven, and the effectiveness of the new algorithm is shown by comparison with other algorithms. [ABSTRACT FROM AUTHOR]

Resource-constrained project scheduling problem (RCPSP) is a kind of representative practical engineering problem, the purpose is to schedule activities in the project through rational use of limited resources in the minimum time. This paper proposes an improved hybrid cuckoo algorithm (CS&SA) to solve the RCPSP problem. Firstly, the individual elements are randomly coded into priority vectors, and the population is decoded using serial scheduling to convert the individual into a set of task scheduling sequences. Secondly, the Levy flight is redesigned to change the algorithm from random walk to adaptive as the population fitness changes. Then, this article adds three neighborhood update techniques to meet the update requirements of the algorithm at different stages. Finally, in order to prevent the algorithm from falling into a local optimum, this paper introduces a simulated annealing strategy to allow the algorithm to accept some individuals with poor quality with a certain probability in each iteration. In the testing part, this paper firstly tests the effectiveness and optimization of three different-scale examples in the classic example library PSPLIB for RCPSP problems. Among them, the average error of the CS&SA algorithm in the small-scale J30 is 0.25%, and the average error in the medium-scale J60 is 11.21%, and the average error of the large-scale J120 is 19.83%. Then, by comparing other intelligent optimization algorithms, it is proved that NCS&SA is superior to other algorithms in optimization and accuracy. [ABSTRACT FROM AUTHOR]

The main purpose of this paper is to determine a non-trivial tractable class of the maximum (k,m)-subsets problem. More specifically, we show that we can solve this problem in polynomial time for m = 1. Depending on the value of k with respect to n, we prove that in the best-case scenario our algorithm runs in O(√n) time. On the other hand, an upper bound for solving it is given by O(n4/k³). Furthermore, for the case where n = k², we design and implement an algorithm in Python that yields a solution in O(n5/2). [ABSTRACT FROM AUTHOR]

This paper deals with multi-objective school bus routing problem, which includes route balance, total number of school buses and total travel distance optimization objectives. An improved non-dominated sorting genetic algorithm (NSGAII) is proposed to solve this problem. First, the definition of measurement indicator that denotes the degree of route balance is given based on the analysis of the solved problem. And then, the multi-objective optimization function is provided. In the proposed algorithm, the individuals are obtained by using the tournament selection, sequential crossover and inverse mutation. The 2-opt neighborhood operator is adopted to improve the best individuals obtained in each iteration. At the same time, the route selection rule based on the degree of route balance is applied to select the final optimal solution set. The solution with better balance degree will be taken as the best solution. Finally, some benchmark instances are used to test the effectiveness of proposed algorithm. The results reveal that the proposed algorithm outperforms the standard NSGAII and Multi-objective Evolutionary Algorithm(MOEA). The experimental results also show that our algorithm has good stability. [ABSTRACT FROM AUTHOR]

The simulated annealing algorithm is a stochastic optimization method for solving optimization problems associated with a large search space. The efficiency of the algorithm depends on the adaptation of the cooling model. Despite this, the main disadvantage of this algorithm is that it does not take into account the state of the system while searching. Thus, it is difficult to predict the system convergence with the simulated annealing algorithm. However, neural networks in particular continuous Hopfield networks have proven their ability in the field of machine learning to make a decision. In this paper, we introduce continuous Hopfield networks to improve the convergence of the simulated annealing algorithm. The experimental results show that the hybrid approach produces a large number of stable sets. [ABSTRACT FROM AUTHOR]

The distribution of power flow can be effectively improved by optimizing the configuration of Distributed Generators (DGs) and Shunt Capacitors (SCs). Proper placements and capacities of DGs or SCs for real power loss minimization at the radial distribution network (RDN) is attracting more attention as people become more and more dependent on electric power resources. In order to overcome the slow convergence and poor effect of intelligent algorithms in this problem, Particle Swarm Optimization with Orientation and Shrinking factor (PSO-OS) is proposed in this paper. In the suggested PSO-OS algorithm, sensitivity analysis sort of distribution network is introduced to guide the direction of variation to simplify the search space, and shrinking factor is introduced to balance search capabilities in various RDN systems. The proposed method is applied to IEEE-33, IEEE-69 and IEEE-119 systems to solve the problem of simultaneous allocation of DGs and SCs. Simulation results show that the energy loss of the system is reduced and the stability is improved after reconfiguration. By repeating the experiment independently with the traditional PSO algorithm, the comparative analysis of the results proves the superior performance of the improved algorithm. By comparing with the results in recent literature, it reveals the effectiveness of the proposed PSO-OS algorithm in finding the optimal position and capacity of DGs and SCs in the distribution network. [ABSTRACT FROM AUTHOR]

Water cycle algorithm (WCA) is a heuristic algorithm proposed in recent years. To overcome the insufficiency of standard WCA algorithm in solving the non-convex optimal power flow (OPF) problems, the multi-objective novel improved water cycle algorithm (MONIWCA) is proposed in this paper. The evaporation process is improved in WCA by introducing evaporation rate and the normal distribution optimization mechanism is used to mutate the individual position. The modified WCA also adopts a constraint-based strategy to ensure zero constraint violations. In order to obtain a high-quality Pareto optimal solution set (POS) and select the best compromise solution (BCs), a global ranking strategy is proposed. The global ranking strategy includes the novel constraint handling method, the rank index calculation and the BCs on fuzzy satisfaction theory to deal with the complex constraints of the optimization problem. The MONIWCA has been simulated under the constraints of zero violations on IEEE 30, IEEE 57 and IEEE 118 standard test systems, including six dual-objective cases and one tri- objective case. The simulation results are compared with the multi-objective particle swarm optimization (MOPSO). The results show that the improved method can effectively solve the MOOPF problem, not only to obtain a uniform continuous Pareto solution set but also to achieve a better compromise solution. In addition, the two performance indicators of the generational distance (GD) and the spacing (SP) also show that the MONIWCA algorithm has uniform distribution, high convergence and strong stability. [ABSTRACT FROM AUTHOR]

It is very difficult to solve indefinite quadratic programming problem. In this paper, for globally solving such problem, a novel algorithm is presented. Firstly, the initial problem (P) is converted into an equivalent problem (EP); then, the problem (EP) is reduced to a sequence of linear programming problems, which are very easy to be solved. The convergence and the complexity of the proposed algorithm is presented, and some experiments are provided to show its feasibility. [ABSTRACT FROM AUTHOR]

For globally solving sum of linear ratios problem (SLRP), this paper presents a new branch-and-bound method. In this method, a new linear relaxation technique is proposed firstly; then, the initial problem SLRP is solved by a sequence of linear programming problems. Meanwhile, to improve the convergence speed of our algorithm, two accelerating techniques are presented. The proposed algorithm is proved to be convergent, and some experiments are provided to show its feasibility and efficiency. [ABSTRACT FROM AUTHOR]

In this paper, for solving generalized linear fractional programming (GLFP), a new branch-and-reduce approach is presented. Firstly, an equivalent problem (EP) of GLFP is given; then, a new linear relaxation technique is proposed; finally, the problem EP is reduced to a sequence of linear programming problems by using the new linear relaxation technique. Meanwhile, to improve the convergence speed of our algorithm, two reducing techniques are presented. The proposed algorithm is proved to be convergent, and some experiments are provided to show its feasibility and efficiency. [ABSTRACT FROM AUTHOR]

In this paper, we consider a generalized longest common subsequence problem, in which a constraining sequence of length s must be included as a substring and the other constraining sequence of length t must be included as a subsequence of two main sequences and the length of the result must be maximal. For the two input sequences X and Y of lengths n and m, and the given two constraining sequences of length s and t, we present an O(nmst) time dynamic programming algorithm for solving the new generalized longest common subsequence problem. The time complexity can be reduced further to cubic time in a more detailed analysis. The correctness of the new algorithm is proved. [ABSTRACT FROM AUTHOR]

Different regularities can be used to identify the sequence among other sequences. Regularities allow us to infer an information about the evolution of the sequence. Tandem repeats are the most frequent in the genomes of eukaryotes. Extraction of regularities is a widely studied problem. However, searching for exact tandem repeats can be too restrictive. So, a natural extension of the repetition is to allow errors. In this paper, we consider the approximate period problem. In particular, we consider an explicit reduction from the approximate period problem to the satisfiability problem and present experimental results for different satisfiability algorithms. Also, we consider the approximate period problem for sequences of motor primitives of robots. In particular, we use the approximate period problem to obtain some meta-parameters that adapt the global motion behavior. We try to use such meta-parameters for learning to generalize motor primitives to a different behavior by trial and error without re-learning the task. [ABSTRACT FROM AUTHOR]

This paper proposes a exact solution algorithm to explicitly obtain the exact optimal solution of a second-order cone programming problem with box constraints of decision variables. The proposed solution algorithm is based on a parametric solution approach to determine the optimal strict region of parameters, and the main procedures are to perform deterministic equivalent transformations for the main problem and to solve the KKT condition of auxiliary problem without the loss of optimality. [ABSTRACT FROM AUTHOR]

This paper describes a method of considering the problem of planning for indoor service robot. We introduce a novel graph theoretic formulation of the planning a typical working day problem. We prove that this problem is NP-complete. In proof we consider an explicit reduction of the planning a typical working day problem to the problem of finding a Hamiltonian path. Also we describe an approach to solve this problem. This approach is based on constructing a logical model for the planning a typical working day problem. Using our reduction from the planning a typical working day problem to the Hamiltonian path problem and reduction from the Hamiltonian path problem to the satisfiability problem one can try to solve the planning a typical working day problem by local search and intelligent algorithms which developed for the satisfiability problem. [ABSTRACT FROM AUTHOR]