The series of studies about the convergence or not of the evolutionary strategies of players that use co-evolutionary genetic algorithms in Cournot games has not addressed the issue of individual players' strategi...
详细信息
ISBN:
(纸本)9783642122385
The series of studies about the convergence or not of the evolutionary strategies of players that use co-evolutionary genetic algorithms in Cournot games has not addressed the issue of individual players' strategies convergence, but only of the convergence of the aggregate indices (total quantity and price) to the levels that correspond either to the Nash or Walrash Equilibrium. Here we discover that while some algorithms lead to convergence of the aggregates to Nash Equilibrium values, this is not the case for the individual players' strategies (i.e. no NE is reached). Co-evolutionary programming social learning, as well as a social learning algorithm we introduce here, achieve this goal (in a stochastic sense);this is displayed by statistical tests, as well as "NE stages" evaluation, based on ergodic Markov chains.
In order to rank the performance of machine learning algorithms, many researchers conduct experiments on benchmark data sets. Since most learning algorithms have domain-specific parameters, it is a popular custom to a...
详细信息
ISBN:
(纸本)1558604804
In order to rank the performance of machine learning algorithms, many researchers conduct experiments on benchmark data sets. Since most learning algorithms have domain-specific parameters, it is a popular custom to adapt these parameters to obtain a minimal error rate on the test set. The same rate is then used to rank the algorithm, which causes an optimistic bias, We quantify this bias, showing, in particular, that an algorithm with more parameters will probably be ranked higher than an equally good algorithm with fewer parameters. We demonstrate this result, showing the number of parameters and trials required in order to pretend to outperform C4.5 or FOIL, respectively, for various benchmark problems. We then describe out how unbiased ranking experiments should be conducted.
An advance in economic thought is in the area of behavioral economics where traditional models of rational decision-making are challenged by newer models of behavior such as Prospect Theory. This is coupled with a wor...
详细信息
ISBN:
(纸本)9781424498642
An advance in economic thought is in the area of behavioral economics where traditional models of rational decision-making are challenged by newer models of behavior such as Prospect Theory. This is coupled with a world where algorithms have abilities to learn, remember and evolve over time to make better decisions. The advances on these two fronts are forcing the world of markets to be analyzed from a different angle. This work is a look at markets to compare traditional expected utility theory of economic decision-making to the newer idea of Prospect Theory. Two learning algorithms, based on traditional expected utility and Prospect Theory, are designed and then compared under several scenarios designed to replicate various market conditions faced by investors. Deviations were analyzed to measure the effectiveness of the two algorithms and also the two models of economic decision making, where it was found that risk averseness described by Prospect Theory will lead to greater deviations in expected prices than more traditional models of economic decision making. This is for several reasons, including risk aversion can, in most situations, lead to suboptimal economic decisions.
A max-min criterion for design of bidirectional associative memory, which requires the smallest domain of attraction to be maximized, is proposed in this paper. A quick learning algorithm is first given, by which the ...
详细信息
A max-min criterion for design of bidirectional associative memory, which requires the smallest domain of attraction to be maximized, is proposed in this paper. A quick learning algorithm is first given, by which the designed connection weights are 1, 0 or-1. Further, a constrained perception optimization algorithm is presented, which takes the weights obtained by quick algorithm as initial iteration value. Computer experimental results confirm the advantages of the proposed algorithms.
The paper presents the application of signal flow graphs (SFG) and adjoint flow graphs (AFG) in determination the gradient vector for feedforward neural networks. The presented approach is universal and applicable in ...
详细信息
ISBN:
(纸本)0818674563
The paper presents the application of signal flow graphs (SFG) and adjoint flow graphs (AFG) in determination the gradient vector for feedforward neural networks. The presented approach is universal and applicable in the same form irrespective of the particular structure of the network. The applicability of the method has been shown on examples of different types of neural networks: multilayer perceptron, sigma-pi network, generalized radial basis network and multilayer Volterra network. The method finds application in any gradient based learning algorithms of neural networks. Some applications of this method, concerning the prediction and identification of the nonlinear dynamic plants are presented and discussed in the paper.
We study the convergence of a class of gradient-based Model-Agnostic Meta-learning (MAML) methods and characterize their overall complexity as well as their best achievable accuracy in terms of gradient norm for nonco...
详细信息
We study the convergence of a class of gradient-based Model-Agnostic Meta-learning (MAML) methods and characterize their overall complexity as well as their best achievable accuracy in terms of gradient norm for nonconvex loss functions. We start with the MAML method and its first-order approximation (FO-MAML) and high-light the challenges that emerge in their analysis. By overcoming these challenges not only we provide the first theoretical guarantees for MAML and FO-MAML in nonconvex settings, but also we answer some of the unanswered questions for the implementation of these algorithms including how to choose their learning rate and the batch size for both tasks and datasets corresponding to tasks. In particular, we show that MAML can find an epsilon-first-order stationary point (epsilon-FOSP) for any positive epsilon after at most O(1/epsilon(2)) iterations at the expense of requiring second-order information. We also show that FO-MAML which ignores the second-order information required in the update of MAML cannot achieve any small desired level of accuracy, i.e., FO-MAML cannot find an epsilon-FOSP for any epsilon > 0. We further propose a new-variant of the MAML algorithm called Hessian-free MAML which preserves all theoretical guarantees of MAML, without requiring access to second-order information.
This work summarizes our research on the topic of the application of unsupervised learning algorithms to the problem of intrusion detection, and in particular our main research results in network intrusion detection. ...
详细信息
ISBN:
(纸本)9781424420650
This work summarizes our research on the topic of the application of unsupervised learning algorithms to the problem of intrusion detection, and in particular our main research results in network intrusion detection. We proposed a novel, two tier architecture for network intrusion detection, capable of clustering packet payloads and correlating anomalies in the packet stream. We show the experiments we conducted on such architecture, we give performance results, and we compare our achievements with other comparable existing systems.
Meta-learning helps us find solutions to computational intelligence (CI) challenges in automated way. Meta-learning algorithm presented in this paper is universal and may be applied to any type of CI problems. The nov...
详细信息
ISBN:
(纸本)9781424418206
Meta-learning helps us find solutions to computational intelligence (CI) challenges in automated way. Meta-learning algorithm presented in this paper is universal and may be applied to any type of CI problems. The novelty of our proposal lies in complexity controlled testing combined with very useful learning machines generators. The simplest and the best solutions are strongly preferred and are explored earlier. The learning algorithm is augmented by meta-knowledge repository which accumulates information about progress of the search through the space of candidate solutions. The approach facilitates using human experts knowledge to restrict the search space and provide goal definition, gaining meta-knowledge in an automated manner.
Radio tomographic imaging (RTI) is an emerging technique which obtains images of passive targets (i.e., not carrying electronic device) within a wireless sensor network using received signal strength (RSS). One major ...
详细信息
Radio tomographic imaging (RTI) is an emerging technique which obtains images of passive targets (i.e., not carrying electronic device) within a wireless sensor network using received signal strength (RSS). One major problem that restricts the application of RTI is the difficulty to model the variations of RSS measurements caused by moving targets in different multi-path environments. This paper proposes to apply background learning algorithm to RTI system to model variations. Compared with previous RSS-based device free localization methods, the proposed method achieves higher accuracy in multi-target and time-varying environment without offline training. Firstly, two fundamental background learning algorithms, mixture of gaussians and kernel density estimation, are introduced to calculate the probabilities of links being affected by targets using RSS measurement. Then, Tikhonov regularization is applied to the reconstruction of images using the probabilities. Experimental results show that the proposed approach achieves high accuracy and increases the RSS-network capacity considerably.
Progress in the field of machine learning has been fueled by the introduction of benchmark datasets pushing the limits of existing algorithms. Enabling the design of datasets to test specific properties and failure mo...
详细信息
ISBN:
(纸本)9781713829546
Progress in the field of machine learning has been fueled by the introduction of benchmark datasets pushing the limits of existing algorithms. Enabling the design of datasets to test specific properties and failure modes of learning algorithms is thus a problem of high interest, as it has a direct impact on innovation in the field. In this sense, we introduce Synbols - Synthetic Symbols - a tool for rapidly generating new datasets with a rich composition of latent features rendered in low resolution images. Synbols leverages the large amount of symbols available in the Unicode standard and the wide range of artistic font provided by the open font community. Our tool's high-level interface provides a language for rapidly generating new distributions on the latent features, including various types of textures and occlusions. To showcase the versatility of Synbols, we use it to dissect the limitations and flaws in standard learning algorithms in various learning setups including supervised learning, active learning, out of distribution generalization, unsupervised representation learning, and object counting.
暂无评论