Lianbo Ma

Artificial Bee Colony Optimizer Based on Bee Life-Cycle for Stationary and Dynamic Optimization

This paper proposes a novel optimization scheme by hybridizing an artificial bee colony optimizer (HABC) with a bee life-cycle mechanism, for both stationary and dynamic optimization problems. The main innovation of the proposed HABC is to develop a cooperative and population-varying scheme, in which individuals can dynamically shift their states of birth, foraging, death, and reproduction throughout the artificial bee colony life cycle. That is, the bee colony size can be adjusted dynamically according to the local fitness landscape during algorithm execution. This new characteristic of HABC helps to avoid redundant search and maintain diversity of population in complex environments. A comprehensive experimental analysis is implemented that the proposed algorithm is benchmarked against several state-of-the-art bio-inspired algorithms on both stationary and dynamic benchmarks. Then the proposed HABC is applied to the real-world applications including data clustering and image segmentation problems. Statistical analysis of all these tests highlights the significant performance improvement due to the life-cycle mechanism and shows that the proposed HABC outperforms the reference algorithms.

Cooperative two-engine multi-objective bee foraging algorithm with reinforcement learning

Abstract This paper proposes a novel multi-objective bee foraging algorithm (MOBFA) based on two-engine co-evolution mechanism for solving multi-objective optimization problems. The proposed MOBFA aims to handle the convergence and diversity separately via evolving two cooperative search engines with different evolution rules. Specifically, in the colony-level interaction, the primary concept is to first assign two different performance evaluation principles (i.e., Pareto-based measure and indicator-based measure) to the two engines for evolving each archive respectively, and then use the comprehensive learning mechanism over the two archives to boost the population diversity. In the individual-level foraging, the neighbor-discount-information (NDI) learning based on reinforcement learning (RL) is integrated into the single-objective searching to adjust the flight trajectories of foraging bee. By testing on a suit of benchmark functions, the proposed MOBFA is verified experimentally to be superior or at least comparable to its competitors in terms of two commonly used metrics IGD and SPREAD.

Two-Level Master-Slave RFID Networks Planning via Hybrid Multiobjective Artificial Bee Colony Optimizer

Radio frequency identification (RFID) networks planning (RNP) is a challenging task on how to deploy RFID readers under certain constraints. Existing RNP models are usually derived from the flat and centralized-processing framework identified by vertical integration within a set of objectives which couple different types of control variables. This paper proposes a two-level RNP model based on the hierarchical decoupling principle to reduce computational complexity, in which the cost-efficient planning at the top levels is modeled with a set of discrete control variables (i.e., switch states of readers), and the quality of service objectives at the bottom level are modeled with a set of continuous control variables (i.e., physical coordinate and radiate power). The model of the objectives at the two levels is essentially a multiobjective problem. In order to optimize this model, this paper proposes a specific multiobjective artificial bee colony optimizer called H-MOABC, which is based on performance indicators with reinforcement learning and orthogonal Latin squares approach. The proposed algorithm proves to be competitive in dealing with two-objective and three-objective optimization problems in comparison with state-of-the-art algorithms. In the experiments, H-MOABC is employed to solve the two scalable real-world RNP instances in the hierarchical decoupling manner. Computational results shows that the proposed H-MOABC is very effective and efficient in RFID networks optimization.

Optimal Controller Placement Problem in Internet-Oriented Software Defined Network

The feasibility, scalability and performance of Internet-oriented Software Defined Network (SDN) are confronted with severe challenge, so it needs multiple controllers that influence every aspect of SDN performance. An effective placement algorithm of these controllers is especially important. In this paper, we formulate the Multiobjective Optimization Controller Placement (MOCP) problem and focus on maximizing network reliability, maximizing controller load balance ability and minimizing latency between controllers and switches and then solve the optimal selection of controller placement locations, within which nodes are under control of each controller and the optimal distribution of their routing requests among these controllers. We formulate this problem into a mathematical model as the optimization objective function. In order to resolve this model, Adaptive Bacterial Foraging Optimization (ABFO) algorithm is developed on account of the computation complexity according to actual network state. The results show this proposed scheme The results show this proposed scheme has great potential to deal with above optimization objectives efficiently and effectively.

Multi-indicator Bacterial Foraging Algorithm with Kriging Model for Many-Objective Optimization

In order to efficiently reduce computational expense as well as manage the diversity and convergence in many-objective optimization, this paper proposes a novel multi-indicator bacterial foraging algorithm with Kriging model (K-MBFA) to guide the search process toward the Pareto front. In the proposed algorithm, a set of preferential individuals for the improved Kriging model are appropriately selected according to the different indicators. Specifically, the stochastic ranking technique is adopted to avoid the search biases of different indicators, which would lead the population to converge to local region of the Pareto front. With several test instances from DTLZ sets with 3, 5, 8 and 10 objectives, K-MBFA is verified to be significantly superior to other compared algorithms in terms of inverted generational distance (IGD).

A Novel Many-Objective Bacterial Foraging Optimizer Based on Multi-engine Cooperation Framework.

In order to efficiently manage the diversity and convergence in many-objective optimization, this paper proposes a novel multi-engine cooperation bacterial foraging algorithm (MCBFA) to enhance the selection pressure towards Pareto front. The main framework of MCBFA is to handle the convergence and diversity separately by evolving several search engines with different rules. In this algorithm, three engines are respectively endowed with three different evolution principles (i.e., Pareto-based, decomposition-based and indicator-based), and their archives are evolved according to comprehensive learning. In the foraging operations, each bacterium is evolved by reinforcement learning (RL). Specifically, each bacterium adaptively varies its own run-length unit and exchange information to dynamically balance exploration and exploitation during the search process. Empirical studies on DTLZ benchmarks show MCBFA exhibits promising performance on complex many-objective problems.

Biomimicry of plant root growth using bioinspired foraging model for data clustering

Clustering is a popular data mining technique widely used in many fields. Recently, researches on swarm intelligence-based and bionic approaches for handing these clustering problems have made significant achievements. In this contribution, a bionic algorithm inspired by the intrinsic adaptability of plant root foraging behavior is designed and developed for data clustering. Especially, the foraging behaviors of plant root involve elongation, branching, and tropism based on the auxin-regulated mechanism. By incorporating the self-adaptive population-varying mechanism and self-adaptive root growth strategy, a new root system growth algorithm with adaptive population variation (RSGA_APV) is designed based on the root foraging and auxin-based regulation of the root system. The comprehensive experimental analysis is implemented that the proposed RSGA_APV is benchmarked against several state-of-the-art reference algorithms on a set of scalable benchmarks. Then, RSGA_APV is applied to resolve data clustering problems. Computational results verify the effectiveness and efficiency of our proposed algorithm.

An adaptive approach for handling two-dimension influence maximization in social networks: Influence Maximization in Social Networks

1College of Computer Science and Engineering, Northeastern University, Shenyang, China 2College of Software, Northeastern University, Shenyang, China 3College of Information Science and Engineering, State Key Laboratory of Synthetical Automation for Process Industries, Northeastern University, Shenyang, China 4College of Information Science and Engineering, Northeastern University, Shenyang, China

Heuristics-based influence maximization for opinion formation in social networks

Abstract Influence maximization for opinion formation (IMOF) is a significant issue in social networks. In general, the main goal of the IMOF is to determine a set of optimal nodes in an effective way for the maximum propagation of the most ideal opinions. The current studies mainly focus on the control methods for the opinion formation, for example, the informed agents, the improved opinion formation models and the effective initial-node-determination algorithms. However, there are few researches on the mathematical modeling of the IMOF. Additionally, the effective mechanisms should be specifically designed to solve the IMOF. In this paper, a general IMOF model is formulated by the informed agents, and then a 3-hop heuristic algorithm is proposed to deal with the IMOF. Firstly, the IMOF is formulated mathematically as an optimization model. Then, in order to satisfy various preferences among nodes, the weighted bounded confidence model is devised to calculate the opinion of each node effectively. Moreover, the influence spread obeys 3-hop rule, and node influence gradually dissipates and stops beyond 3 hop. Therefore, based on node influence, a 3-hop heuristic algorithm is proposed to effectively determine the top-m influential nodes. With the opinion set initialized, a discount strategy is presented to accelerate the propagation of the selected top-m influential nodes. Finally, the experimental results demonstrate that the proposed method obtains better performance on the average of opinions than the chosen benchmarks.

Root system growth biomimicry for global optimization models and emergent behaviors

Terrestrial plants have evolved remarkable adaptability that enables them to sense environmental stimuli and use this information as a basis for governing their growth orientation and root system development. In this paper, we explain the foraging behaviors of plant root and develop simulation models based on the principles of adaptation processes that view root growing as optimization. This provides us with new methods for global optimization. Accordingly a novel bioinspired optimizer, namely the root system growth algorithm (RSGA), is proposed, which adopts the root foraging, memory and communication and auxin-regulated mechanism of the root system. Then RSGA is benchmarked against several state-of-the-art reference algorithms on a suit of CEC2014 functions. Experimental results show that RSGA can obtain satisfactory performances on several benchmarks in terms of accuracy, robustness and convergence speed. Moreover, a comprehensive simulation is conducted to investigate the explicit adaptability of root system in RSGA. That is, in order to be able to climb noisy gradients in nutrients in soil, the foraging behaviors of root system are social and cooperative that is analogous to animal foraging behaviors.