The accurate classification of protein structure is critical and essential for protein function determination in Bioinformatics and Proteomics. A reasonably high rate of prediction accuracy for protein structure class...
详细信息
The accurate classification of protein structure is critical and essential for protein function determination in Bioinformatics and Proteomics. A reasonably high rate of prediction accuracy for protein structure classification has been achieved recently in coarse-grained protein class assignment according to their primary amino acid sequences, such as classifying proteins into four classes in SCOP. However, it is still quite a challenge for fine-grained protein fold assignment, especially when the number of possible folding patterns as those defined in SCOP is large. In our previous work, hierarchical learning architecture (HLA) neural networks have been used to differentiate proteins according to their classes and folding patterns. A better prediction accuracy rate for 27 folding categories was 65.5% which improves previous results by Ding and Dubchak with 56.5% prediction accuracy rate. The success of the protein structure classification depends heavily on the computational methods used and the features selected. Here combinatorial fusion analysis (CFA) techniques are used to facilitate feature selection and combination for improving prediction accuracy rate of protein structure classification. The resulting classification has an overall prediction accuracy rate of 87.8% for coarse-grained 4 classes and 70.9% for fine-grained 27 folding categories by applying the concept of CFA to our previous work using neural network with the HLA framework. These results are significantly higher than others and our previous work. They further demonstrate that the CFA techniques can greatly enhance the machine learning method (such as NN in the paper) in the protein structure prediction problem.
Active Queue Management (AQM) applies a suitable control policy upon detecting congestion in networks. in this paper, an adaptive Proportional-Integral (PI) controller based on Artificial Neural Networks (ANN) is appl...
详细信息
Robust nonlinear controller design with constraint on the poles' location of the linear part of closed-loop system is proposed. The design method is based on the integrator backstepping procedure and linear constr...
详细信息
This paper addresses the design of state feedback H∞ robust controller that satisfies additional constraints on closed-loop poles location. Desired closed-loop poles location is considered to be a clipped sector, cor...
详细信息
The problem of designing a linear controller for nonlinear polynomial systems that results the largest domain of attraction is considered. Moreover, using the proposed algorithm, the shape of the approximation of doma...
详细信息
The objective of this research is to investigate the utility of micro genetic algorithms (GA) to determine the initial conditions of a spacecraft at the point of departure from Earth's orbit to make a successful l...
详细信息
The objective of this research is to investigate the utility of micro genetic algorithms (GA) to determine the initial conditions of a spacecraft at the point of departure from Earth's orbit to make a successful lunar landing or a lunar flyby. The states of the spacecraft at the point at the destination are then obtained using the patched-conic approximation. As demonstrated in this work, faster convergence is obtained using micro genetic algorithms. This methodology can also be extended to interplanetary translation by utilizing the intermediate thrust of the spacecraft.
In this paper, an intelligent controller is applied to speed control of a switched reluctance motor. Two techniques are used which have been successfully used in other intelligent modeling and control applications. Fi...
详细信息
In this paper, fuzzy sliding mode controller is designed to govern the dynamics of switched reluctance motor. First, magnetic characteristics data of the motor obtained by Finite Element Method (FEM) is utilized for a...
详细信息
In this paper, robust regulation of a class of nonlinear singularly perturbed systems, via nonlinear H ∞ approach is considered. Under appropriate assumptions, it is shown through two new theorems that the existence ...
详细信息
In this paper, robust regulation of a class of nonlinear singularly perturbed systems, via nonlinear H ∞ approach is considered. Under appropriate assumptions, it is shown through two new theorems that the existence of a positive definite solution for the Hamilton-Jacobi-Isaacs inequality related to the problem of disturbance attenuation for the main singularly perturbed system, can be related to the existence of a solution of a (simpler) Hamilton-Jacobi-Isaacs inequality arising in the problem of disturbance attenuation for the reduced-order system.
Adaptive Constant False Alarm Rate (CFAR) detectors operating on envelope-squared matched-filtered beam data generally employ two equal fixed-length windows, straddling the test cell, to collect reference samples. Our...
详细信息
ISBN:
(纸本)0933957343
Adaptive Constant False Alarm Rate (CFAR) detectors operating on envelope-squared matched-filtered beam data generally employ two equal fixed-length windows, straddling the test cell, to collect reference samples. Our objective is to prevent the consideration of reference sample sets that do not behave even approximately i.i.d. A procedure to adaptively determine window sizes is presented. The nested hypothesis testing procedure is based on the principles of Quality control and utilizes the Mann-Kendall rank test for randomness. Potential benefits include improved false alarm control, increased detection probability and closer tracking of extended clutter edges.
暂无评论