A Modified Back Propagation Algorithm of Neural Network with Global Optimization

2014 ◽  
Vol 1042 ◽  
pp. 232-238
Author(s):  
Jing Jing Li ◽  
Zhe Cui
Keyword(s):  
Neural Network ◽  
Global Optimization ◽  
Optimization Algorithm ◽  
Back Propagation ◽  
Bp Algorithm ◽  
Efficient Global Optimization ◽  
Back Propagation Algorithm ◽  
Global Optimization Algorithm ◽  
Function Fitting ◽  
Neural Network Algorithm

The advantages and weakens of traditional BP algorithm is briefly analyzed and an efficient global optimization algorithm is proposed.The basic principle of the algorithm is presented,and a new BP neural network algorithm based on the existing BP algorithm and the new global optimization algorithm is proposed, considering the new global optimization algorithm can solve the problem of local minimum efficiently. To verify the effectiveness of the new BP algorithm,the paper compared the experimental results of various algorithms in solving function fitting problem.

scholarly journals A New Hybrid Grasshopper Optimization - Backpropagation for Feedforward Neural Network Training

2020 ◽  
Vol 25 (1) ◽  
pp. 118
Author(s):  
Samer Alsammarraie ◽  
Nazar K. Hussein
Keyword(s):  
Neural Network ◽  
Optimization Algorithm ◽  
Hybrid Algorithm ◽  
Back Propagation ◽  
Global Optimum ◽  
Bp Algorithm ◽  
Back Propagation Algorithm ◽  
Grasshopper Optimization Algorithm ◽  
Grasshopper Optimization ◽  
Searching Ability

The Grasshopper optimization algorithm showed a rapid converge in the initial phases of the global search, however while being around the global optimum, the searching process became so slow. On the contrary, the gradient descending method around achieved faster convergent speed global optimum, and the convergent accuracy was showed to be higher at the same time. As a result, the proposed hybrid algorithm combined Grasshopper optimization algorithm (GOA) along with the back-propagation (BP) algorithm, also referred to as GOA–BP algorithm, was introduced to provide training to the weights of the feed forward neural network (FNN), the proposed hybrid algorithm can utilize the strong global searching ability of the GOA, and the intense local searching ability of the Back-Propagation algorithm. The results of experiments showed that the proposed hybrid GOA–BP algorithm was better and faster in convergent speed and accuracy than the Grasshopper optimization algorithm (GOA) and BP algorithm.   http://dx.doi.org/10.25130/tjps.25.2020.018  

A time simulated annealing-back propagation algorithm and its application in disease prediction

2018 ◽  
Vol 32 (25) ◽  
pp. 1850303 ◽  
Author(s):  
Fang Hu ◽  
Mingzhu Wang ◽  
Yanhui Zhu ◽  
Jia Liu ◽  
Yalin Jia
Keyword(s):  
Neural Network ◽  
Simulated Annealing ◽  
Bp Neural Network ◽  
Back Propagation ◽  
Bp Algorithm ◽  
Disease Prediction ◽  
Back Propagation Algorithm ◽  
Network Algorithm ◽  
Learning Rates ◽  
Neural Network Algorithm

In this paper, based on the Back Propagation (BP) neural network algorithm, we introduce the idea of the Simulated Annealing (SA), and then propose a new neural network algorithm: Time Simulated Annealing-Back Propagation (TSA-BP) algorithm. The proposed algorithm can improve the convergence rate and numerical stability. By using this proposed algorithm, the learning rates and initial weights in the BP neural network could be easily adjusted. We show that the TSA-BP algorithm could reduce the errors caused by human-made factors. Several numerical experiments have been tested by using different disease data. Furthermore, we compared the TSA-BP algorithm to the other existing, well-known algorithms. Numerical results show higher accuracy and efficiency of the TSA-BP algorithm.

scholarly journals Topological Optimization of Under-Platform Dampers With Moving Morphable Components and Global Optimization Algorithm for Nonlinear Frequency Response

2021 ◽  
Vol 143 (2) ◽  
Author(s):  
E. Denimal ◽  
F. El Haddad ◽  
C. Wong ◽  
L. Salles
Keyword(s):  
Global Optimization ◽  
Optimization Algorithm ◽  
Nonlinear Behavior ◽  
Topological Optimization ◽  
Efficient Global Optimization ◽  
Nonlinear Frequency ◽  
Global Optimization Algorithm ◽  
Nonlinear Frequency Response ◽  
The Impact ◽  
Moving Morphable Components

Abstract To limit the risk of high cycle fatigue, underplatform dampers (UDPs) are traditionally used in aircraft engines to control the level of vibration. Many studies demonstrate the impact of the geometry of the damper on its efficiency, thus the consideration of topological optimization (TO) to find the best layout of the damper seems natural. Because of the nonlinear behavior of the structure due to the friction contact interface, classical methods of TO are not usable. This study proposes to optimize the layout of an UDP to reduce the level of nonlinear vibrations computed with the multiharmonic balance method (MHBM). The approach of TO employed is based on the moving morphable components (MMC) framework together with the Kriging and the efficient global optimization algorithm to solve the optimization problem. The results show that the level of vibration of the structure can be reduced to 30% and allow for the identification of different efficient geometries.

Designing a Morphable Parabolic Reflector Antenna Using Origami-Inspired Discretization and Efficient Global Optimization

2020 ◽  
Author(s):  
Gregory Wilson ◽  
Dimitris Lagoudas ◽  
Darren Hartl
Keyword(s):  
United States ◽  
Global Optimization ◽  
Shape Memory Alloy ◽  
Shape Memory ◽  
Radiation Pattern ◽  
Optimization Algorithm ◽  
The United States ◽  
Lessons Learned ◽  
Efficient Global Optimization ◽  
Global Optimization Algorithm

Abstract This research explores a segmented parabolic antenna that can change its physical shape via shape memory alloy actuators, thereby altering its radiation pattern when transmitting a signal. The parabolic dish has been discretized into an origami pattern to make use of the naturally compliant fold regions, about which shape memory alloy wires create moments. Modeling of antenna deformation is accomplished via Abaqus considering SMA wires contracting due to temperature change as a manifestation of the shape memory effect. An electromagnetic analysis of the deformed antenna follows in ANSYS-HFSS to determine the antenna gain in all directions around the structure. The computed radiation pattern is projected onto a goal shape (e.g. the contiguous United States) to determine the degree to which the shaped broadcast pattern matches that of a desired broadcast area. Finally, the design is iterated using an efficient global optimization algorithm to ascertain an actuation schedule that generates the most conformal broadcast pattern. Traditional optimization algorithms such as genetic or particle swarm may require thousands of designs, particularly when many design variables are considered. The efficient global optimization algorithm employs far fewer designs by fitting surrogate models to the data and only testing points where large improvement is expected, thus reducing design optimization time. The evolution and improvement to an antenna will be discussed for an antenna making use of eight, 16, and 24 SMA linear actuators to most optimally broadcast to only the United States while avoiding signal spill-over into other regions, and the lessons learned can then applied to match broadcast pattern based on other countries as well.

Sign in / Sign up

Export Citation Format

Share Document