One of the major challenges facing Artificial Intelligence in the future is the design of trustworthy algorithms. The development of trustworthy algorithms will be a key challenge in Artificial Intelligence for years ...
详细信息
One of the major challenges facing Artificial Intelligence in the future is the design of trustworthy algorithms. The development of trustworthy algorithms will be a key challenge in Artificial Intelligence for years to come. cultural algorithms (CAs) are viewed as one framework that can be employed to produce a trustable evolutionary algorithm. They contain features to support both sustainable and explainable computation that satisfy requirements for trustworthy algorithms proposed by Cox [Nine experts on the single biggest obstacle facing AI and algorithms in the next five years, Emerging Tech Brew, January 22, 2021]. Here, two different configurations of CAs are described and compared in terms of their ability to support sustainable solutions over the complete range of dynamic environments, from static to linear to nonlinear and finally chaotic. The Wisdom of the Crowds method was selected for the one configuration since it has been observed to work in both simple and complex environments and requires little long-term memory. The Common Value Auction (CVA) configuration was selected to represent those mechanisms that were more data centric and required more long-term memory content. Both approaches were found to provide sustainable performance across all the dynamic environments tested from static to chaotic. Based upon the information collected in the Belief Space, they produced this behavior in different ways. First, the topologies that they employed differed in terms of the "in degree" for different complexities. The CVA approach tended to favor reduced "indegree/outdegree", while the WM exhibited a higher indegree/outdegree in the best topology for a given environment. These differences reflected the fact the CVA had more information available for the agents about the network in the Belief Space, whereas the agents in the WM had access to less available knowledge. It therefore needed to spread the knowledge that it currently had more widely throughout the pop
This study focuses on utilizing neuroevolution for the task of pathology detection in chest radiographs using the CheXpert dataset. The principal objective will be to adapt and extend the TensorFlow-NeuroEvolution (tf...
详细信息
ISBN:
(纸本)9798350385359
This study focuses on utilizing neuroevolution for the task of pathology detection in chest radiographs using the CheXpert dataset. The principal objective will be to adapt and extend the TensorFlow-NeuroEvolution (tfne) framework to the CheXpert dataset, dense blocks, and cultural algorithms. This involves the integration of a dense block module that emulates the dense blocks and transition layers found in DenseNet as well as an attention mechanism, into the tfne framework. Like the original DenseNet architecture, this module will include the definition of a dense block, and a transition layer. Unlike the original DenseNet architecture, the dense block module will include an attention mechanism. Further, a custom CheXpert environment will be created to train the evolved networks on the task of chest pathology detection. Utilizing the CoDeep- NEAT algorithm and cultural algorithms, the study will evolve a neural network equipped for handling the CheXpert dataset using the dense block module. cultural algorithms will help guide the CoDeepNEAT algorithm by providing belief space, and influencing the mutation, crossover, and selection processes. The belief space will also serve as a method of maintaining diversity in the population.
The goal of this paper is to investigate the applicability of evolutionary algorithms to the design of real-time industrial controllers. Present-day "deep learning" (DL) is firmly established as a useful too...
详细信息
The goal of this paper is to investigate the applicability of evolutionary algorithms to the design of real-time industrial controllers. Present-day "deep learning" (DL) is firmly established as a useful tool for addressing many practical problems. This has spurred the development of neural architecture search (NAS) methods in order to automate the model search activity. CATNeuro is a NAS algorithm based on the graph evolution concept devised by Neuroevolution of Augmenting Topologies (NEAT) but propelled by cultural algorithm (CA) as the evolutionary driver. The CA is a network-based, stochastic optimization framework inspired by problem solving in human cultures. Knowledge distribution (KD) across the network of graph models is a key to problem solving success in CAT systems. Two alternative mechanisms for KD across the network are employed. One supports cooperation (CATNeuro) in the network and the other competition (WM). To test the viability of each configuration prior to use in the industrial setting, they were applied to the design of a real-time controller for a two-dimensional fighting game. While both were able to beat the AI program that came with the fighting game, the cooperative method performed statistically better. As a result, it was used to track the motion of a trailer (in lateral and vertical directions) using a camera mounted on the tractor vehicle towing the trailer. In this second real-time application (trailer motion), the CATNeuro configuration was compared to the original NEAT (elitist) method of evolution. CATNeuro is found to perform statistically better than NEAT in many aspects of the design including model training loss, model parameter size, and overall model structure consistency. In both scenarios, the performance improvements were attributed to the increased model diversity due to the interaction of CA knowledge sources both cooperatively and competitively.
This paper compares several approaches to cooperative multi-agent path planning (MAP) based upon variations of the A* algorithm. To simulate multi-agent migration patterns three path-finding mechanisms based on the cl...
详细信息
ISBN:
(数字)9781665467087
ISBN:
(纸本)9781665467087
This paper compares several approaches to cooperative multi-agent path planning (MAP) based upon variations of the A* algorithm. To simulate multi-agent migration patterns three path-finding mechanisms based on the classic A* algorithm was utilized: A*;A*, Ambush and Dendriform. Each makes different assumptions about group leadership in terms of their path generation. A* assumes a single leader for the migratory group: A*mbush allows the group to move in waves;and Dendriform allows the group to decompose and recompose into groups of arbitrary size with local leaders. Each mechanism required parameter weightings so that the simulated agents would interact realistically with their environment. cultural algorithms were employed to adjust the parameter weight categories in order to optimize the group movement under each of these leadership strategies. The three approaches were applied to the simulation of a real-world multi-agent system, the migration of large herd of caribou. The simulated migration was part of the Deepdive Virtual Reality system. In those simulations A* with a single planning agent emphasized nutrition at the expense of the other parameters. A*mbush learned to reduce nutrition slightly and while increasing its emphasis on risk and exploration. On the other hand, Dendriform emphasized overall effort since its planning more dynamic and required more concentration on local effort to be optimized.
cultural algorithms (CAs) are evolutionary algorithms (EAs) inspired by the conceptual models of the human cultural evolution process. In contrast to the conventional EAs, which work only based on the population space...
详细信息
cultural algorithms (CAs) are evolutionary algorithms (EAs) inspired by the conceptual models of the human cultural evolution process. In contrast to the conventional EAs, which work only based on the population space, CAs employ an additional space, called belief space, to collect the information about the behaviour of individuals in the search space. Since the emergence of CAs, they have been successfully extended to solve a wide variety of problems in different branches of science and technology. In this paper, a comprehensive survey on the recent advances in CAs is presented. Literature survey reveals some interesting challenges and future research directions.
cultural algorithms have led to the development of many ways to distribute information within social networks. These mechanisms act by helping the system make decisions about how information is distributed through a p...
详细信息
ISBN:
(数字)9781728169293
ISBN:
(纸本)9781728169293
cultural algorithms have led to the development of many ways to distribute information within social networks. These mechanisms act by helping the system make decisions about how information is distributed through a population network, and thus are called distribution or decision mechanisms. Many distribution mechanisms have been developed using techniques from auction theory, game theory and various forms of voting construct. Here we discuss several methods of Knowledge distribution collectively called the auction distributions mechanisms and their performance is compared using dynamic complex real-valued functional landscapes. We perform this comparison with regards to robustness, how well the system finds solutions, and resilience, how well the system reacts to changes in the dynamics of the system. In this paper an additional Subcultured Distribution Mechanism is described that works to factor the knowledge distribution mechanism into subnetworks in order to support a "deep social learning" approach. The Subcultured Distribution Mechanism is compared with the results of each individual distribution mechanism without a subculture enhancement, when applied to a series of dynamic complex optimization problems of varying complexities. The results suggest that relatively simple mechanism such as Weighted Majority Wins and First Price Auction are sufficient for environments that exhibit low entropic levels of change such as in linear changing environments. Fur non linearly changing environments, First Price Multi-round and English Auctions are most of effective on their own. The Subcultured Distribution Mechanism extension of these mechanisms was found to he best suited for complexities where the two distribution mechanisms had similar performances, and in the most chaotic environments where having multiple distribution mechanisms to choose from was advantageous.
In Computation intelligence algorithm performance is crucial especially when the complexity of the system increases and becomes chaotic (unpredictable). In cultural Systems many algorithms are able to predict the syst...
详细信息
ISBN:
(纸本)9781728187082
In Computation intelligence algorithm performance is crucial especially when the complexity of the system increases and becomes chaotic (unpredictable). In cultural Systems many algorithms are able to predict the system performance as the complexity is linear, or non-linear. However, when it is chaotic the prediction quality decreases dramatically. In this paper, we are show that Common Value Auctions are able to distribute sufficient information through the system in order to sustain the prediction rate even on the edge of chaos. This sustainability is expressed here in terms of increased resilience and robustness. Systems that rely on wisdom of the crowd based approaches are shown not to do as well when environmental change goes from linear to non-linear, and finally to chaotic.
Recently it has been found that the earth's oceans are warming at a pace that is 40% faster than predicted by a United Nations panel a few years ago. As a result, 2018 has become the warmest year on record for the...
详细信息
ISBN:
(纸本)9781728169293
Recently it has been found that the earth's oceans are warming at a pace that is 40% faster than predicted by a United Nations panel a few years ago. As a result, 2018 has become the warmest year on record for the earth's oceans. That is because the oceans have acted as a buffer by absorbing 93% of the heat produced by the greenhouse gases [1]. The impact of the oceanic warming has already been felt in terms of the periodic warming of the Pacific Ocean as an effect of the ENSO process. The ENSO process is a cycle of warming and subsequent cooling of the Pacific Ocean that can last over a period of years. This cycle was first documented by Peruvian fishermen in the early 1600's. So it has been part of the environmental challenges that have been presented to economic agents throughout the world since then. It has even been suggested that the cycle has increased in frequency over the years, perhaps in response to the overall issues related to global warming. [2] [3] In this paper cultural algorithms are used to develop a multi-objective agent-based model of artisanal (traditional offshore) fishing behavior in coastal Peru, Cerro Azul. The data used to produce this model comes from the observation of fishing behavior over a four year period, 1982-1986. During this period over 6000 individual fishing trips were documented. This observation period overlapped with one of the largest ENSO activities ever recorded. As a result, it was possible to observe the changes in fishing behavior that were the result of this process. While the data is several decades old, the ENSO process was first observed in Peru in 1502. Thus, this data can be considered to reflect the adaptations that have been made to the process in the ensuing centuries. The model was used to produce Pareto curves that reflected tradeoffs in terms of fish quality and trip effort during each of three phases on the ENSO process. A version of cultural algorithms, CAPSO, was then used to compute whether these curves
In cultural Systems there are many ways to collect and distribute problem solving knowledge within social networks. Such mechanisms include games, auctions, and various voting mechanisms. Here, a new distribution mech...
详细信息
ISBN:
(纸本)9781728155845
In cultural Systems there are many ways to collect and distribute problem solving knowledge within social networks. Such mechanisms include games, auctions, and various voting mechanisms. Here, a new distribution mechanism, Common Value Auctions, is presented. In this paper Common Value Auctions are used to distribute problem solving knowledge within a social network The experimental results suggest that the Common Value Auction distribution mechanism is sufficient to produce sustainable learning in a dynamically changing problem landscape using only information about an individuals' past influence by a knowledge agent. Here, the sustainability is expressed in terms of both the cultural Systems' robustness and resilience to linear changes in the problem solving environment. Future work will investigate whether additional variables are required to produce sustainability in non-linear and chaotic environments.
Workflow scheduling has remained a critical functionality of modern data-centric workflow management systems. Cloud computing, which provides practically unlimited computing and storage resources, has enabled a new ge...
详细信息
ISBN:
(纸本)9781728100593
Workflow scheduling has remained a critical functionality of modern data-centric workflow management systems. Cloud computing, which provides practically unlimited computing and storage resources, has enabled a new generation of data-centric workflows, called big data workflows. New big data workflow scheduling algorithms should optimally utilize the characteristics of cloud computing such as heterogeneous virtual machines, the elastic resource provisioning model, and the pay-as-you-go pricing model, as well as the time and monetary cost to transfer large amounts of data. In this paper, we consider one case of the general big data workflow scheduling problem where a deadline, 6, is given for a workflow, W, and the goal is to minimize the monetary cost of running W in the cloud while satisfying the given deadline, 6. To this end, we leverage the power of Evolutionary algorithms (EA) in order to search for the best solution within a reasonable planning time. More specifically, we introduce an innovative fitness function that combines the time and monetary cost of a workflow in one metric. Based on the EA and the fitness function, we design a deadline-constrained big data workflow scheduling algorithm, called iCATS (Improved cultural algorithms-based Task Scheduling). Extensive experiments demonstrate the statistical advantages of iCATS over existing representative EA workflow scheduling algorithms, including random (Rand), Genetic algorithms (GA), Particle Swarm Optimization (PSO), and cultural algorithms (CA).
暂无评论