Hybrid ant colony algorithm for texture classification

We present a novel ant colony algorithm integrating genetic algorithms and simplex algorithms. This method is able to not only speed up searching process for optimal solutions, but also improve the quality of the solutions. The proposed method is applied to set up a learning model for the "tuned" mask, which is used for texture classification. Experimental results on real world images and comparisons with genetic algorithms and genetic simplex algorithms are presented to illustrate the merit and feasibility of the proposed method.

A multi-objective ant colony optimization algorithm based on the physarum-inspired mathematical model

Multi-objective traveling salesman problem (MOTSP) is an important field in operations research, which has wide applications in the real world. Multi-objective ant colony optimization (MOACO) as one of the most effective algorithms has gained popularity for solving a MOTSP. However, there exists the problem of premature convergence in most of MOACO algorithms. With this observation in mind, an improved multiobjective network ant colony optimization, denoted as PMMONACO, is proposed, which employs the unique feature of critical tubes reserved in the network evolution process of the Physarum-inspired mathematical model (PMM). By considering both pheromones deposited by ants and flowing in the Physarum network, PM-MONACO uses an optimized pheromone matrix updating strategy. Experimental results in benchmark networks show that PM-MONACO can achieve a better compromise solution than the original MOACO algorithm for solving MOTSPs.

Localized ant colony of robots for redeployment in wireless sensor networks

Sensor failures or oversupply in wireless sensor networks (WSNs), especially initial random deployment, create spare sensors (whose area is fully covered by other sensors) and sensing holes. We envision a team of robots to relocate sensors and improve their area coverage. Existing algorithms, including centralized ones and the only localized G-R3S2 [9], move only spare sensors and have limited improvement because non-spare sensors, with area coverage mostly overlapped by neighbour sensors, are not moved, and additional sensors are deployed to fill existing small holes. We propose a localized algorithm, called Localized Ant-based Sensor Relocation Algorithm with Greedy Walk (LASR-G), where each robot may carry at most one sensor and makes decision that depends only on locally detected information. In LASRG, each robot calculates corresponding pickup or dropping probability, and relocates sensor with currently low coverage contribution to another location where sensing hole would be significantly reduced. The basic algorithm optimizes only area coverage, while modified algorithm includes also the cost of robot movement. We compare LASR-G with G-R3S2, and examine both single robot and multi robots scenarios. The simulation results show the advantages of LASR-G over G-R3S2.

An ant colony system based on the physarum network

The Physarum Network model exhibits the feature of important pipelines being reserved with the evolution of network during the process of solving a maze problem. Drawing on this feature, an Ant Colony System (ACS), denoted as PNACS, is proposed based on the Physarum Network (PN). When updating pheromone matrix, we should update both pheromone trails released by ants and the pheromones flowing in a network. This hybrid algorithm can overcome the low convergence rate and local optimal solution of ACS when solving the Traveling Salesman Problem (TSP). Some experiments in synthetic and benchmark networks show that the efficiency of PNACS is higher than that of ACS. More important, PNACS has strong robustness that is very useful for solving a higher dimension TSP.

Bayesian networks and decision trees in the diagnosis of female urinary incontinence

This study compares the effectiveness of Bayesian networks versus Decision Trees in modeling the Integral Theory of Female Urinary Incontinence diagnostic algorithm. Bayesian networks and Decision Trees were developed and trained using data from 58 adult women presenting with urinary incontinence symptoms. A Bayesian Network was developed in collaboration with an expert specialist who regularly utilizes a non-automated diagnostic algorithm in clinical practice. The original Bayesian network was later refined using a more connected approach. Diagnoses determined from all automated approaches were compared with the diagnoses of a single human expert. In most cases, Bayesian networks were found to be at least as accurate as the Decision Tree approach. The refined Connected Bayesian Network was found to be more accurate than the Original Bayesian Network accurately discriminated between diagnoses despite the small sample size. In contrast, the Connected and Decision Tree approaches were less able to discriminate between diagnoses. The Original Bayesian Network was found to provide an excellent basis for graphically communicating the correlation between symptoms and laxity defects in a given anatomical zone. Performance measures in both networks indicate that Bayesian networks could provide a potentially useful tool in the management of female pelvic floor dysfunction. Before the technique can be utilized in practice, well-established learning algorithms should be applied to improve network structure. A larger training data set should also improve network accuracy, sensitivity, and specificity.

Learning Bayesian networks in Semi-deterministic systems

In current constraint-based (Pearl-style) systems for discovering Bayesian networks, inputs with deterministic relations are prohibited. This restricts the applicability of these systems. In this paper, we formalize a sufficient condition under which Bayesian networks can be recovered even with deterministic relations. The sufficient condition leads to an improvement to Pearl’s IC algorithm; other constraint-based algorithms can be similarly improved. The new algorithm, assuming the sufficient condition proposed, is able to recover Bayesian networks with deterministic relations, and moreover suffers no loss of performance when applied to nondeterministic Bayesian networks.

A universal optimization strategy for ant colony optimization algorithms based on the Physarum-inspired mathematical model

Ant colony optimization (ACO) algorithms often fall into the local optimal solution and have lower search efficiency for solving the travelling salesman problem (TSP). According to these shortcomings, this paper proposes a universal optimization strategy for updating the pheromone matrix in the ACO algorithms. The new optimization strategy takes advantages of the unique feature of critical paths reserved in the process of evolving adaptive networks of the Physarum-inspired mathematical model (PMM). The optimized algorithms, denoted as PMACO algorithms, can enhance the amount of pheromone in the critical paths and promote the exploitation of the optimal solution. Experimental results in synthetic and real networks show that the PMACO algorithms are more efficient and robust than the traditional ACO algorithms, which are adaptable to solve the TSP with single or multiple objectives. Meanwhile, we further analyse the influence of parameters on the performance of the PMACO algorithms. Based on these analyses, the best values of these parameters are worked out for the TSP.

A novel Physarum-Based ant colony system for solving the real-world traveling salesman problem

The solutions to Traveling Salesman Problem can be widely applied in many real-world problems. Ant colony optimization algorithms can provide an approximate solution to a Traveling Salesman Problem. However, most ant colony optimization algorithms suffer premature convergence and low convergence rate. With these observations in mind, a novel ant colony system is proposed, which employs the unique feature of critical tubes reserved in the Physaurm-inspired mathematical model. A series of experiments are conducted, which are consolidated by two realworld Traveling Salesman Problems. The experimental results show that the proposed new ant colony system outperforms classical ant colony system, genetic algorithm, and particle swarm optimization algorithm in efficiency and robustness.

A rank-based ant system algorithm for solving 0/1 knapsack problem

The 0/1 Knapsack Problem (KP), which is a classical NP-complete problem, has been widely applied to solving many real world problems. Ant system (AS), as one of the earliest ant colony optimization (ACO) algorithms, provides approximate solutions to 0/1 KPs. However, there are some shortcomings such as low efficiency and premature convergence in most AS algorithms. In order to overcome the shortcomings of AS, this paper proposes a rank-based AS algorithm, denoted as RAS to solve 0/1 KP. Taking advantages of the ranked ants with a higher profit, the pheromone of items will be updated with better solutions in RAS. Experimental results in different datasets show that this new kind of AS algorithm can obtain a higher efficiency and robustness when solving 0/1 KP.

The application of ant colony systems to image texture classification

This paper presents a novel ant system based optimisation method which integrates genetic algorithms and simplex algorithms. This method is able to not only speed up the search process for solutions, but also improve the quality of the solutions. In this paper, the proposed method is applied to set up a learning model for the "tuned" mask, which is used for texture classification. Experimental results on aerial images and comparisons with genetic algorithms and genetic simplex algorithms are presented to illustrate the merit and feasibility of the proposed method.

Layered dynamic probabilistic networks for spatio-temporal modelling

In applications such as tracking and surveillance in large spatial environments, there is a need for representing dynamic and noisy data and at the same time dealing with them at different levels of detail. In the spatial domain, there has been work dealing with these two issues separately, however, there is no existing common framework for dealing with both of them. In this paper, we propose a new representation framework called the Layered Dynamic Probabilistic Network (LDPN), a special type of Dynamic Probabilistic Network (DPN), capable of handling uncertainty and representing spatial data at various levels of detail. The framework is thus particularly suited to applications in wide-area environments which are characterised by large region size, complex spatial layout and multiple sensors/cameras. For example, a building has three levels: entry/exit to the building, entry/exit between rooms and moving within rooms. To avoid the problem of a relatively large state space associated with a large spatial environment, the LDPN explicitly encodes the hierarchy of connected spatial locations, making it scalable to the size of the environment being modelled. There are three main advantages of the LDPN. First, the reduction in state space makes it suitable for dealing with wide area surveillance involving multiple sensors. Second, it offers a hierarchy of intervals for indexing temporal data. Lastly, the explicit representation of intermediate sub-goals allows for the extension of the framework to easily represent group interactions by allowing coupling between sub-goal layers of different individuals or objects. We describe an adaptation of the likelihood sampling inference scheme for the LDPN, and illustrate its use in a hypothetical surveillance scenario.

An amoeboid algorithm for shortest path in fuzzy weighted networks

Taking the uncertainty existing in edge weights of networks into consideration, finding shortest path in such fuzzy weighted networks has been widely studied in various practical applications. In this paper, an amoeboid algorithm is proposed, combing fuzzy sets theory with a path finding model inspired by an amoeboid organism, Physarum polycephalum. With the help of fuzzy numbers, uncertainty is well represented and handled in our algorithm. What's more, biological intelligence of Physarum polycephalum has been incorporate into the algorithm. A numerical example on a transportation network is demonstrated to show the efficiency and flexibility of our proposed amoeboid algorithm.