In this paper, a framework for replacing missing values in a database is proposed since a real-world database is seldom complete. Good data quality in a database can directly improve the performance of any data mining...
详细信息
In this paper, a framework for replacing missing values in a database is proposed since a real-world database is seldom complete. Good data quality in a database can directly improve the performance of any data mining algorithm in various applications. Our proposed framework adopts the basic concepts from conditional probability theories and further develops an algorithm to facilitate the capability of handling both nominal and numerical values, which addresses the problem of the inability of handling both nominal and numerical values with a high degree of accuracy in the existing algorithms. Several experiments are conducted and the experimental results demonstrate that our framework provides a high accuracy when compared with most of the commonly used algorithms such as using the average value, using the maximum value, and using the minimum value to replace missing values.
This paper is concerned with the structural stability of boundary equilibria in a class of hybrid systems, that of piecewise linear continuous systems. Specifically, we study the structural stability under parameter v...
详细信息
This paper is concerned with the structural stability of boundary equilibria in a class of hybrid systems, that of piecewise linear continuous systems. Specifically, we study the structural stability under parameter variations of equilibria lying on discontinuity boundaries in phase space dividing regions where the system under investigation is smooth. We show that it is possible to give a set of conditions to account for the possible dynamical scenarios that can be observed. We present a novel scenario where under perturbations, the equilibrium of interest does not persist and a family of stable limit cycles is generated with amplitudes increasing linearly under parameter perturbations. We propose a novel switching control strategy based on the use of the results presented in the paper.
Artificial neural networks have been traditionally employed to learn and compute the inverse kinematics of a robotic arm. However, the inverse kinematics model of a typical robotic arm with joint limits is a multi-val...
详细信息
Artificial neural networks have been traditionally employed to learn and compute the inverse kinematics of a robotic arm. However, the inverse kinematics model of a typical robotic arm with joint limits is a multi-valued and discontinuous function. Because it is difficult for a multilayer neural network to approximate this type of function, an accurate inverse kinematics model cannot be obtained by using a single neural network. In order to overcome the difficulties of inverse kinematics learning, we propose a novel modular neural network system that consists of a number of expert modules, where each expert approximates a continuous part of the inverse kinematics function. The proposed system selects one appropriate expert whose output minimizes the expected position/orientation error of the end-effector of the arm. The system can learn a precise inverse kinematics model of a robotic arm with equal or more degrees of freedom than that of its end-effector. However, there are robotic arms with fewer degrees of freedom, where the system cannot learn their precise inverse kinematics model. We have adopted a modified Gauss-Newton method for finding the least-squares solution to address this issue. Through the modifications presented in this paper, the improved modular neural network system can obtain a precise inverse kinematics model of a general robotic arm.
Repetitive processes are a distinct class of 2D systems (i.e. information propagation in two independent directions) of both systems theoretic and applications interest. They cannot be controlled by direct extension o...
详细信息
Repetitive processes are a distinct class of 2D systems (i.e. information propagation in two independent directions) of both systems theoretic and applications interest. They cannot be controlled by direct extension of existing techniques from either standard (termed 1D here) or 2D systems theory. Here we give the first results on how feedback/feedforward control action can be used to influence one form of controllability for processes with discrete dynamics.
In this paper, electrical characteristics of germanium (Ge) nanowire field effect transistors (FETs) are computationally investigated. A calibrated three-dimensional (3D) density-gradient simulation is performed to ex...
详细信息
ISBN:
(纸本)4990276205
In this paper, electrical characteristics of germanium (Ge) nanowire field effect transistors (FETs) are computationally investigated. A calibrated three-dimensional (3D) density-gradient simulation is performed to explore the electrical characteristics of Ge omega-shaped-gate FETs. The examined nanodevices are with a 70% gate coverage ratio. By evaluating the threshold voltage roll-off, the transfer characteristics, and the leakage current, our numerical results have shown that the Ge nanowire FET has potentially higher driving-capability than that of the silicon (Si) one. Due to good channel controllability of the omega-shaped-gate FET with the 70% gate coverage ratio, compared with the Si nanowire FET, the high mobility Ge nanowire FET significantly suppresses the effect of band-gap narrowing on the transport characteristics. Leakage current of the Ge nanowire FET depends upon the thickness of the gate channel film. A thinner Ge film leads to a lower leakage current. Preliminary numerical study on the Ge Nanowire FET with a propoer selection on gate material provides interesting results for the design and fabrication of high performance nanodevices in nanoelectronics era.
Resource management for individual flows can significantly improve quality of service (QoS) in mobile cellular networks. However, its efficiency depends on the availability of information about the movement of mobile ...
详细信息
ISBN:
(纸本)0780382773
Resource management for individual flows can significantly improve quality of service (QoS) in mobile cellular networks. However, its efficiency depends on the availability of information about the movement of mobile terminals. Movement prediction can potentially provide this information, but is costly if performed by the network and usually assumes a certain movement model, which may not adequately reflect each individual user's behavior. Instead, we propose the concept of movement contracts, where a mobile terminal specifies its movement to the network, which in return provides a better QoS as long as the provided specification is sufficiently accurate. We describe approaches to specify the spatial and temporal aspects of movement with a parsimonious parameter set and evaluate these approaches through simulations. We find that movement contracts can significantly reduce both session blocking and handover dropping probability simultaneously, whereas existing approaches have to make a tradeoff between these.
暂无评论