Deep multilayer multiple kernel learning

Multiple kernel learning (MKL) approach has been proposed for kernel methods and has shown high performance for solving some real-world applications. It consists on learning the optimal kernel from one layer of multiple predefined kernels. Unfortunately, this approach is not rich enough to solve relatively complex problems. With the emergence and the success of the deep learning concept, multilayer of multiple kernel learning (MLMKL) methods were inspired by the idea of deep architecture. They are introduced in order to improve the conventional MKL methods. Such architectures tend to learn deep kernel machines by exploring the combinations of multiple kernels in a multilayer structure. However, existing MLMKL methods often have trouble with the optimization of the network for two or more layers. Additionally, they do not always outperform the simplest method of combining multiple kernels (i.e., MKL). In order to improve the effectiveness of MKL approaches, we introduce, in this paper, a novel backpropagation MLMKL framework. Specifically, we propose to optimize the network over an adaptive backpropagation algorithm. We use the gradient ascent method instead of dual objective function, or the estimation of the leave-one-out error. We test our proposed method through a large set of experiments on a variety of benchmark data sets. We have successfully optimized the system over many layers. Empirical results over an extensive set of experiments show that our algorithm achieves high performance compared to the traditional MKL approach and existing MLMKL methods.     

A fast and noise resilient cluster-based anomaly detection

Clustering, while systematically applied in anomaly detection, has a direct impact on the accuracy of the detection methods. Existing cluster-based anomaly detection methods are mainly based on spherical shape clustering. In this paper, we focus on arbitrary shape clustering methods to increase the accuracy of the anomaly detection. However, since the main drawback of arbitrary shape clustering is its high memory complexity, we propose to summarize clusters first. For this, we design an algorithm, called Summarization based on Gaussian Mixture Model (SGMM), to summarize clusters and represent them as Gaussian Mixture Models (GMMs). After GMMs are constructed, incoming new samples are presented to the GMMs, and their membership values are calculated, based on which the new samples are labeled as “normal” or “anomaly.” Additionally, to address the issue of noise in the data, instead of labeling samples individually, they are clustered first, and then each cluster is labeled collectively. For this, we present a new approach, called Collective Probabilistic Anomaly Detection (CPAD), in which, the distance of the incoming new samples and the existing SGMMs is calculated, and then the new cluster is labeled the same as of the closest cluster. To measure the distance of two GMM-based clusters, we propose a modified version of the Kullback–Libner measure. We run several experiments to evaluate the performances of the proposed SGMM and CPAD methods and compare them against some of the well-known algorithms including ABACUS, local outlier factor (LOF), and one-class support vector machine (SVM). The performance of SGMM is compared with ABACUS using Dunn and DB metrics, and the results indicate that the SGMM performs superior in terms of summarizing clusters. Moreover, the proposed CPAD method is compared with the LOF and one-class SVM considering the performance criteria of (a) false alarm rate, (b) detection rate, and (c) memory efficiency. The experimental results show that the CPAD method is noise resilient, memory efficient, and its accuracy is higher than the other methods.     

Structural Properties of Networked Control Systems With Bandwidth Limitations and Delays

The structural properties of networked control systems with both bandwidth limitations and delays are investigated. Sufficient conditions are given for controllability (stabilizability) and reconstructibility (detectability). Our results enhance previous works by capturing bandwidth limitations and delays simultaneously. The adopted modeling framework could be readily used in control and estimation methods, including optimal and predictive schemes. It also facilitates the use of scheduling optimization algorithms in conjunction with the control scheme presented.     

Impacts in multibody systems: modeling and experiments

This paper outlines a novel approach to the modeling and analysis of impact involving multibody systems. This approach is based on an analysis of energy absorption and restitution during impact, using a decomposition of the kinetic energy, which decouples the parts associated with the spaces of admissible and constrained motions of the underlying unilateral constraints. Such a decomposition turns out to be useful in the analysis of energy dissipation during impact, and leads to a generalized definition of the energetic coefficient of restitution, which targets particularly collisions in multibody systems. The applicability of the approach reported is investigated by conducting an experimental study on a robotic testbed. It is shown that impact between multibody systems is considerably affected not only by the local dynamics characteristics of the interacting bodies, but also the configuration of the whole multibody system. The results reported here show that our decomposition can offer a sound characterization of impact in several problems of multibody systems.     

Determining the Main Factors in Declining the Urmia Lake Level by Using System Dynamics Modeling

Urmia Lake in Iran is the second largest saline lake in the world. This ecosystem is the home for different species. Due to various socio-economical and ecological criteria, Urmia Lake has important role in the Northwestern part of the country but it has faced many problems in recent years. Because of droughts, overuse of surface water resources and dam constructions, water level has decreased in such a way that one quarter of the lake has changed to saline area in the last 10 years. The purpose of this research is to determine the main factors which reduce the lake’s water level. To this end, a simulation model, based on system dynamics method, is developed for the Urmia Lake basin to estimate the lake’s level. After successful verification of the model, results show that (among the proposed factors) changes in inflows due to the climate change and overuse of surface water resources is the main factor for 65% of the effect, constructing four dams is responsible for 25% of the problem, and less precipitation on lake has 10% effect on decreasing the lake’s level in the recent years. In the future, the model also can be used by managers as a decision support system to find the effects of building new dams or other infrastructures.     

Multi-product multi-chance-constraint stochastic inventory control problem with dynamic demand and partial back-ordering: A harmony search algorithm

In this paper, a multiproduct inventory control problem is considered in which the periods between two replenishments of the products are assumed independent random variables, and increasing and decreasing functions are assumed to model the dynamic demands of each product. Furthermore, the quantities of the orders are assumed integer-type, space and budget are constraints, the service-level is a chance-constraint, and that the partial back-ordering policy is taken into account for the shortages. The costs of the problem are holding, purchasing, and shortage. We show the model of this problem is an integer nonlinear programming type and to solve it, a harmony search approach is used. At the end, three numerical examples of different sizes are given to demonstrate the applicability of the proposed methodology in real world inventory control problems, to validate the results obtained, and to compare its performances with the ones of both a genetic and a particle swarm optimization algorithms.     

Optimal pricing, lot-sizing and marketing planning in a capacitated and imperfect production system

In the classical economic production quantity (EPQ) problem demand is considered to be known in advance. However, in the real-world, demand of a product is a function of factors such as product’s price, its quality, and marketing expenditures for promoting the product. Quality level of the product and specifications of the adopted manufacturing process also affect the unit product’s cost. Therefore, in this paper we consider a profit maximizing firm who wants to jointly determine the optimal lot-sizing, pricing, and marketing decisions along with manufacturing requirements in terms of flexibility and reliability of the process. Geometric programming (GP) technique is proposed to address the resulting nonlinear optimization problem. Using recent advances in optimization techniques we are able to optimally solve the developed, highly nonlinear, mathematical model. Finally, using numerical examples, we illustrate the solution approach and analyze the solution under different conditions.     

Adaptive Control for State Synchronization of Nonlinear Haptic Telerobotic Systems with Asymmetric Varying Time Delays

In this paper, we introduce a new adaptive controller design scheme for nonlinear telerobotic systems with varying time delays where the delays and their variation rates are unknown. The designed controller has the ability to synchronize the state behaviors of the local and the remote robots. In this paper, asymptotic stability in the presence of varying time delays is of interest. Using the proposed controller, asymptotic stability of the bilateral telerobotic system subject to any bounded yet unknown varying delay with a bounded yet unknown rate of change can be guaranteed. Besides the varying time delay, the proposed adaptive controller has the ability to adapt to the parameter variations in the local and the remote robots’ dynamics. It is shown that position and velocity errors between the local and the remote manipulators converge to the zero asymptotically, thus ensuring teleoperation transparency. Experimental and simulation results with a pair of PHANToM haptic devices and a pair of planar manipulators under varying time delays in the communication channel demonstrate the effectiveness of the proposed scheme.     

Implementation of genetic algorithm in an embedded microcontroller-based polarization control system

Implementation of genetic algorithm in a PIC32MX microcontroller-based polarization control system is proposed and demonstrated. The controller measures the signal intensity that is used to estimate the genetic value. This process is controlled by the genetic algorithm rather than referring to the Look-Up-Table as implemented in existing solutions. To reach optimum performance, the code is optimized by using the best genetic parameters so that the fastest execution time can be achieved. An ability of genetic algorithm to work efficiently in polarization control system possesses many advantages including easy code construction, low memory consumption and fast control speed. Genetic algorithm is intelligent enough to be used for endless polarization stabilization and in the worst case, able to stabilize the polarization changes in 300 μs. In the best case the response time can reach 17 μs.