A Hybrid Ant Colony Algorithm with a Local Search for the Strongly Correlated Knapsack Problem

Author(s):  
Wiem Zouari ◽  
Ines Alaya ◽  
Moncef Tagina
2014 ◽  
Vol 704 ◽  
pp. 257-260
Author(s):  
De Wen Cai ◽  
Chen Fei Shao ◽  
Di Kai Wang ◽  
Er Feng Zhao ◽  
Meng Yang

Back Propagation (BP) neural network can learn and store a large number of input-output model nonlinear relationships with simple structure. Niche ant colony algorithm (NACA) combines the ant colony algorithm (ACA) with the niche technology in order to add its local search ability to ACA with preserving the intelligent search ability and robustness of ACA. To optimize predicting model establishment of the dam monitoring data, NACA and BP neural network modeling method are combined to establish a prediction model of horizontal displacement monitoring data. The traditional BP neural network prediction model is established to make a comparison with the NACA. The results show that NACA-BP neural network method can speed up the convergence rate of BP neural network and enhance local search ability and prediction accuracy.


2013 ◽  
Vol 765-767 ◽  
pp. 658-661
Author(s):  
Yan Zhang ◽  
Hui Ling Wang ◽  
Xu Li ◽  
Yong Hua Zhang ◽  
Hao Wang

To overcome the limitation of precocity and stagnation in classical ant colony algorithm, this article presents a Parallel Ant System Based on OpenMP. The ant colony is divided into three children ant colonies according to the characteristics of natural ant colony multi-group and pheromone updating features of ant colony algorithm. By Open Multi-Processing parallel programming idea, the parallel and cooperating optimization of children ant colonies was obtained. It organically combines local search and global search, makes full use of computing power of multi-core CPU, and improves the efficiency significantly. Contrastive experiments show that the algorithm has a better capability of global optimization than traditional ant colony algorithm.


2011 ◽  
Vol 230-232 ◽  
pp. 973-977 ◽  
Author(s):  
Zhi Jun Hu ◽  
Rong Li

0-1 knapsack problem is a typical combinatorial optimization question in the design and analysis of algorithms. The mathematical description of the knapsack problem is given in theory. The 0-1 knapsack problem is solved by ant colony optimistic algorithm that is improved by introducing genetic operators. To solve the 0-1 knapsack problem with the improved ant colony algorithm, experimental results of numerical simulations, compared with greedy algorithm and dynamic programming algorithm, have shown obvious advantages in efficiency and accuracy on the knapsack problem.


2013 ◽  
Vol 380-384 ◽  
pp. 1877-1880 ◽  
Author(s):  
Rui Tao Liu ◽  
Xiu Jian Lv

This paper uses MapReduce parallel programming mode to make the Ant Colony Optimization (ACO) algorithm parallel and bring forward the MapReduce-based improved ACO for Multi-dimensional Knapsack Problem (MKP). A variety of techniques, such as change the probability calculation of the timing, roulette, crossover and mutation, are applied for improving the drawback of the ACO and complexity of the algorithm is greatly reduced. It is applied to distributed parallel as to solve the large-scale MKP in cloud computing. Simulation experimental results show that the algorithm can improve the defects of long search time for ant colony algorithm and the processing power for large-scale problems.


Sign in / Sign up

Export Citation Format

Share Document