Using spotted hyena optimizer for training feedforward neural networks

Spotted hyena optimizer (SHO) is a novel metaheuristic optimization algorithm based on the behavior of spotted hyena and their collaborative behavior in nature. In this paper, we design a spotted hyena optimizer for training feedforward neural network (FNN), which is regarded as a challenging task since it is easy to fall into local optima. Our objective is to apply metaheuristic optimization algorithm to tackle this problem better than the mathematical and deterministic methods. In order to confirm that using SHO to train FNN is more effective, five classification datasets and three function-approximations are applied to benchmark the performance of the proposed method. The experimental results show that the proposed SHO algorithm for optimization FNN has the best comprehensive performance and has more outstanding performance than other the state-of-the-art metaheuristic algorithms in terms of the performance measures.     

A Classification Framework using a Diverse Intensified Strawberry Optimized Neural Network (DISON) for Clinical Decision-making

A novel classification framework for clinical decision making that uses an Extremely Randomized Tree (ERT) based feature selection and a Diverse Intensified Strawberry Optimized Neural network (DISON) is proposed. DISON is a Feed Forward Artificial Neural Network where the optimization of weights and bias is done using a two phase training strategy. Two algorithms namely Strawberry Plant Optimization (SPO) algorithm and Gradient-descent Back-propagation algorithm are used sequentially to identify the optimum weights and bias. The novel two phase training method and the stochastic duplicate-elimination strategy of SPO helps in addressing the issue of local optima associated with conventional neural networks. The relevant attributes are selected based on the feature importance values computed using an ERT classifier.Vertebral Column, Pima Indian diabetes (PID), Cleveland Heart disease (CHD) and Statlog Heart disease (SHD) datasets from the University of California Irvine machine learning repository are used for experimentation. The framework has achieved an accuracy of 87.17% for Vertebral Column, 90.92% for PID, 93.67% for CHD and 94.5% for SHD. The classifier performance has been compared with existing works and is found to be competitive in terms of accuracy, sensitivity and specificity. Wilcoxon test confirms the statistical superiority of the proposed method.     

Time horizons in rats foraging for food in temporally separated patches

An important tenet of optimal foraging theory is that foragers compare prey densities in alternative patches to determine an optimal distribution of foraging behavior over time. A critical question is over what time period (time horizon) this integration of information and behavior occurs. Recent research has indicated that rats do not compare food density in a depleting patch with that in a rich patch delayed by an hour or more (Timberlake, 1984). In the present research we attempted to specify over what time period a future rich patch would affect current foraging. The effect of future food was measured by early entry into the rich patch (anticipation) and by a decrease in food obtained in the depleting patch (suppression). The rats showed anticipation of a rich patch up to an hour distant, but suppressed current feeding only if the rich patch was 16 min distant or less. The suppression effect appeared mediated by competition for expression between anticipatory entries into the rich patch and continued foraging in the depleting patch. These results suggest that optimal foraging is based on a variety of specific mechanisms rather than a general optimizing algorithm with a single time horizon.     

Research and application of virtual user context information security strategy based on group intelligent computing

This article first introduced the current technology of the privacy protection model, and analyzed their characteristics and deficiencies. Afterwards, from the point of view of revenue, the shortcomings of the traditional privacy protection model have analyzed through the group intelligent computing method. Based on this, this paper proposes a research and application of virtual user information of security strategy based on group intelligent computing, through the collection of visitor's private information historical access data, intelligent calculation of the strategy group between the visitor and the interviewee. The setting of the threshold of the visited person can protect the privacy information of the user more effectively. In this paper, the implementation flow, algorithm implementation process, and specific architecture design of the proposed virtual user of privacy protection model based on group intelligent computing are introduced respectively. The specific algorithms include PCA, BP neural network, and genetic algorithm. Finally, the proposed privacy has verified through experiments. Protection model can protect user privacy more effectively than traditional privacy protection model. In the future, we will further expand and improve the privacy protection model of virtual users based on group intelligent computing, including considering the dynamic and inconsistency of access to the privacy information, that is, accessing different private information will produce different overlay effects and parallelism. We will also study how to apply this model to actual systems such as shopping websites and social platforms, and use commercial data to evaluate the performance of the model and further improve it.     

An improved learning algorithm for random neural networks based on particle swarm optimization and input-to-output sensitivity

Extreme learning machine (ELM) for random single-hidden-layer feedforward neural networks (RSLFN) has been widely applied in many fields in the past ten years because of its fast learning speed and good generalization performance. But because traditional ELM randomly selects the input weights and hidden biases, it typically requires high number of hidden neurons and thus decreases its convergence performance. It is necessary to select optimal input weights and hidden biases to improve the convergence performance of the traditional ELM. Generally, the single-hidden-layer feedforward neural networks (SLFN) with low input-to-output sensitivity will cause good robustness of the network, which may further lead into good generalization performance. Moreover, particle swarm optimization (PSO) has no complicated evolutionary operators and fewer parameters need to adjust, and is easy to implement. In this study, an improved ELM based on PSO and input-to-output sensitivity information is proposed to improve RSLFN’s convergence performance. In the improved ELM, PSO encoding the input to output sensitivity information of the SLFN is used to optimize the input weights and hidden biases. The improved ELM could obtain better generalization performance as well as improve the conditioning of the SLFN by decreasing the input-to-output sensitivity of the network. Finally, experiment results on the regression and classification problems verify the improved performance the proposed ELM.     

A modified conjugate gradient-based Elman neural network

Elman recurrent network is a representative model with feedback mechanism. Although gradient descent method has been widely used to train Elman network, it frequently leads to slow convergence. According to optimization theory, conjugate gradient method is an alternative strategy in searching the descent direction during training. In this paper, an efficient conjugate gradient method has been presented to reach the optimal solution in two ways: (1) constructing a more effective conjugate coefficient, (2) determining adaptive learning rates in terms of the generalized Armijo search method. Experiments show that the performance of the new algorithm is superior to traditional algorithms, such as gradient descent method and conjugate gradient method. In particular, the new algorithm has better performance than the evolutionary algorithm. Finally, we prove the weak and strong convergence of the presented algorithm, i.e., the gradient norm of the error function with respect to the weight vectors converges to zero and the weight sequence approaches a fixed optimal point.     

The contribution of private and public information in foraging by Australasian gannets

Predators that forage on foods with temporally and spatially patchy distributions may rely on private or public sources of information to enhance their chances of foraging success. Using GPS tracking, field observations, and videography, we examined potential sites and mechanisms of information acquisition in departures for foraging trips by colonially breeding Australasian gannets (Morus serrator). Analyses of the bill-fencing ceremony between mated pairs of breeding gannets did not detect correlations between parameters of this reciprocal behavior and foraging trips, as would have been predicted if gannets used this behavior as a source of private information. Instead, 60 % of the departing birds flew directly to join water rafts of other conspecific en route to the feeding grounds. The departure of solitary birds from the water rafts was synchronized (within 60 s) with the arrival of incoming foragers and also among departing birds. Furthermore, solitary departing birds from the rafts left in the same directional quadrant (90º slices) as the prior arriving (67 %) and also prior departing forager (79 %). When associated plunge dives of conspecific were visible from the colony, providing a public source of information, gannets more often departed from the water rafts in groups. Our study thus provides evidence for the use of water rafts, but not the nest site, as locations of information transfer, and also confirms the use of local enhancement as a strategy for foraging flights by Australasian gannets.     

Common marmosets (<Emphasis Type="Italic">Callithrix jacchus</Emphasis>) do not utilize social information in three simultaneous social foraging tasks

Social foraging is suggested to increase foraging efficiency, as individuals might benefit from public information acquired by monitoring the foraging activities of other group members. We conducted a series experiments with captive common marmosets (Callithrix jacchus) to investigate to what extent marmosets utilize social information about food location when foraging simultaneously with conspecifics. Subjects were confronted with dominant and subordinate demonstrators in three experiments which differed in the amount of information about food location available to the demonstrators. In all three experiments, the focal subjects’ performance in the social condition was not enhanced in comparison to a non-social control condition. Because we could rule out kleptoparasitism and aggressive displacements as explanations, we argue that the subjects’ tendency for scramble competition by avoiding others and dispersing over the foraging area seems to inhibit or mask the acquisition of social information about the location of rewarded patches.     

Output‐sensitive complexity of multiobjective combinatorial optimization

We study output‐sensitive algorithms and complexity for multiobjective combinatorial optimization problems. In this computational complexity framework, an algorithm for a general enumeration problem is regarded efficient if it is output‐sensitive, that is, its running time is bounded by a polynomial in the input and the output size. We provide both practical examples of multiobjective combinatorial optimization problems for which such an efficient algorithm exists as well as problems for which no efficient algorithm exists under mild complexity theoretic assumptions.     

临床外伤性视神经病变过度医疗的探讨

外伤性视神经病变是一类严重损害视功能的疾病,临床治疗上普遍存在着过度治疗的问题。原因是多样的,治疗标准不统一是主要原因。依据循证医学的成果,对临床外伤性视神经病变现阶段存在的过度医疗进行分析探讨,并提出了医者针对外伤性视神经病变过度医疗应采取的措施,获得最优化的诊疗方案。     

Choice, changeover, and travel

Since foraging in nature can be viewed as instrumental behavior, choice between sources of food, known as “patches,” can be viewed as choice between instrumental response alternatives. Whereas the travel required to change alternatives deters changeover in nature, the changeover delay (COD) usually deters changeover in the laboratory. In this experiment, pigeons were exposed to laboratory choice situations, concurrent variable-interval schedules, that were standard except for the introduction of a travel requirement for changeover. As the travel requirement increased, rate of changeover decreased and preference for a favored alternative strengthened. When the travel requirement was small, the relations between choice and relative reinforcement revealed the usual tendencies toward matching and undermatching. When the travel requirement was large, strong overmatching occurred. These results, together with those from experiments in which changeover was deterred by punishment or a fixed-ratio requirement, deviate from the matching law, even when a correction is made for cost of changeover. If one accepted an argument that the COD is analogous to travel, the results suggest that the norm in choice relations would be overmatching. This overmatching, however, might only be the sign of an underlying strategy approximating optimization.     

Learning and foraging efficiency in German cockroaches, Blattella germanica (L.) (Insecta: Dictyoptera)

We analysed, under laboratory test conditions, how German cockroach larvae oriented their outgoing foraging trip from their shelter. Our results stressed the importance of external factors, like availability and spatial distribution of food sources, in the choice of a foraging strategy within their home range. When food sources were randomly distributed, larvae adopted a random food search strategy. When food distribution was spatially predictable and reliable, cockroaches were able to relate the presence of food with a landmark during a 3-day training period and to develop an oriented search strategy. Cockroaches were able to associate learned spatial information about their home range to the presence of food resources and then to improve their foraging efficiency. However, conflict experiments revealed that detection of food odour overrode learned landmark cues. Received: 16 October 1999 / Accepted after revision: 18 July 2000     

GROUP FORAGING SENSITIVITY TO PREDICTABLE AND UNPREDICTABLE CHANGES IN FOOD DISTRIBUTION: PAST EXPERIENCE OR PRESENT CIRCUMSTANCES?

The ideal free distribution theory (Fretwell & Lucas, 1970) predicts that the ratio of foragers at two patches will equal the ratio of food resources obtained at the two patches. The theory assumes that foragers have "perfect knowledge" of patch profitability and that patch choice maximizes fitness. How foragers assess patch profitability has been debated extensively. One assessment strategy may be the use of past experience with a patch. Under stable environmental conditions, this strategy enhances fitness. However, in a highly unpredictable environment, past experience may provide inaccurate information about current conditions. Thus, in a nonstable environment, a strategy that allows rapid adjustment to present circumstances may be more beneficial. Evidence for this type of strategy has been found in individual choice. In the present experiments, a flock of pigeons foraged at two patches for food items and demonstrated results similar to those found in individual choice. Experiment 1 utilized predictable and unpredictable sequences of resource ratios presented across days or within a single session. Current foraging decisions depended on past experience, but that dependence diminished when the current foraging environment became more unpredictable. Experiment 2 repeated Experiment I with a different flock of pigeons under more controlled circumstances in an indoor coop and produced similar results.     

Dynamic averaging and foraging decisions in horses (Equus callabus)

The variability of most environments taxes foraging decisions by increasing the uncertainty of the information available. One solution to the problem is to use dynamic averaging, as do some granivores and carnivores. Arguably, the same strategy could be useful for grazing herbivores, even though their food renews and is more homogeneously distributed. Horses (Equus callabus) were given choices between variable patches after short or long delays. When patch information was current, horses returned to the patch that was recently best, whereas those without current information matched choices to the long-term average values of the patches. These results demonstrate that a grazing species uses dynamic averaging and indicate that, like granivores and carnivores, they can use temporal weighting to optimize foraging decisions.     

Majorization as a tool for optimizing a class of matrix functions

The problem of minimizing a general matrix, trace function, possibly subject to certain constraints, is approached by means of majorizing this function by one having a simple quadratic shape and whose minimum is easily found. It is shown that the parameter set that minimizes the majorizing function also decreases the matrix trace function, which in turn provides a monotonically convergent algorithm for minimizing the matrix trace function iteratively. Three algorithms based on majorization for solving certain least squares problems are shown to be special cases. In addition, by means of several examples, it is noted how algorithms may be provided for a wide class of statistical optimization tasks for which no satisfactory algorithms seem available.The Netherlands organization for scientific research(NWO) is gratefully acknowledged for funding this project. This research was conducted while the author was supported by a PSYCHON-grant (560-267-011) from this organization. The author is obliged to Jos ten Berge, Willem Heiser, and Wim Krijnen for helpful comments on an earlier version of this paper.     

Use of visual, acoustic, and olfactory information during embedded invertebrate foraging in brown capuchins (Cebus apella)

Experiments were conducted to investigate which sensory cues are used by brown capuchins (Cebus apella) in embedded invertebrate foraging. The importance of visual, olfactory, and acoustic cues in such foraging was determined by presenting subjects with a stimulus log modified to block out given sensory cues. Experiment 1 was designed to investigate whether subjects could locate an invertebrate embedded in wood when only visual, acoustic, or olfactory information was available. Experiments 2 and 3 were designed to investigate extractive foraging behavior when two sensory cues were provided. It was hypothesized that the combination of visual and acoustic information would be necessary for subjects to successfully locate embedded invertebrates. Results indicated that subjects' performance was most successful when both visual and acoustic information was available.     

The relations between knowledge,search strategy,and performance in unaided and aided information search

Information search is a critical step in resolving complex issues in many different decision-making domains. This study examines the relations between knowledge, search strategy, and performance in both unaided and aided information search. The results of an experiment indicates that (a) task-relevant knowledge is directly related to performance in aided but not in unaided information search; (b) in unaided information search, search strategy mediates the relation between knowledge and performance; that is, knowledge indirectly affects performance through its effect on search strategy; and (c) an information search aid moderates both the relation between knowledge and search strategy and the relation between search strategy and performance. These findings highlight the importance and the roles of task-relevant technical knowledge and search strategy in explaining information search performance and are most likely to apply to complex, knowledge-intensive domains in which decision makers search large databases to resolve issues.     

An improved stochastic EM algorithm for large-scale full-information item factor analysis

In this paper, we explore the use of the stochastic EM algorithm (Celeux & Diebolt (1985) Computational Statistics Quarterly, 2, 73) for large-scale full-information item factor analysis. Innovations have been made on its implementation, including an adaptive-rejection-based Gibbs sampler for the stochastic E step, a proximal gradient descent algorithm for the optimization in the M step, and diagnostic procedures for determining the burn-in size and the stopping of the algorithm. These developments are based on the theoretical results of Nielsen (2000, Bernoulli, 6, 457), as well as advanced sampling and optimization techniques. The proposed algorithm is computationally efficient and virtually tuning-free, making it scalable to large-scale data with many latent traits (e.g. more than five latent traits) and easy to use for practitioners. Standard errors of parameter estimation are also obtained based on the missing-information identity (Louis, 1982, Journal of the Royal Statistical Society, Series B, 44, 226). The performance of the algorithm is evaluated through simulation studies and an application to the analysis of the IPIP-NEO personality inventory. Extensions of the proposed algorithm to other latent variable models are discussed.     

Honeybee methodology, cognition, and theory: recording local directional decisions in interpatch foraging and interpreting their theoretical relevance

Investigations made into the cognitive decision making of honeybees (Apis mellifera) traveling from one flower patch to another flower patch (interpatch foraging) are few. To facilitate such research, we present methods to artificially emulate interpatch foraging and quantify the immediate decision making of honeybees (within 50 cm) choosing to fly an interpatch path. These “Interpatch Methods” are validated, applied, and shown to produce novel information for the field of honeybee spatial cognition. Generally, we demonstrate that a single foraging cohort of honeybees is shown to be capable of making decisions based upon different sets of learned cues, in the exact same context. Specifically, both terminal beacon orientation cues and compass navigation cues can guide the cognitive decision making of interpatch foraging honeybees; our bees chose both cues equally. Finally, the theoretical importance of decision making for interpatch paths is compared with the other foraging paths (outward and homing) with respect to the information available to recruited foragers and scout foragers. We conclude that the ability to analyze interpatch foraging is critical for a more complete understanding of honeybee foraging cognition and that our methods are capable of providing that understanding.     

Monitoring traffic in future cities with aerial swarms: Developing and optimizing a behavior-based surveillance algorithm

Traffic monitoring is a key issue to develop smarter and more sustainable cities in the future, allowing to make a better use of the public space and reducing pollution. This work presents an aerial swarm that continuously monitors the traffic in SwarmCity, a simulated city developed in Unity game engine where drones and cars are modeled in a realistic way. The control algorithm of the aerial swarm is based on six behaviors with twenty-three parameters that must be tuned. The optimization of parameters is carried out with a genetic algorithm in a simplified and faster simulator. The best resulting configurations are tested in SwarmCity showing good efficiencies in terms of observed cars over total cars during time windows. The algorithm reaches a good performance making use of an acceptable computational time for the optimization.