A Resource Limited Immune Approach for Evolving Architecture and Weights of Multilayer Neural Network
نویسندگان
چکیده
A resource limited immune approach (RLIA) was developed to evolve architecture and initial connection weights of multilayer neural networks. Then, with Back-Propagation (BP) algorithm, the appropriate connection weights can be found. The RLIA-BP classifier, which is derived from hybrid algorithm mentioned above, is demonstrated on SPOT multi-spectral image data, vowel data and Iris data effectively. The simulation results demonstrate that RLIA-BP classifier possesses better performance comparing with Bayes maximum-likelihood classifier, k-nearest neighbor classifier (k-NN), BP neural network (BP-MLP) classifier and Resource limited artificial immune classifier (AIRS) in pattern classification.
منابع مشابه
Neural Network Sensitivity to Inputs and Weights and its Application to Functional Identification of Robotics Manipulators
Neural networks are applied to the system identification problems using adaptive algorithms for either parameter or functional estimation of dynamic systems. In this paper the neural networks' sensitivity to input values and connections' weights, is studied. The Reduction-Sigmoid-Amplification (RSA) neurons are introduced and four different models of neural network architecture are proposed and...
متن کاملAdaptive RBF network control for robot manipulators
TThe uncertainty estimation and compensation are challenging problems for the robust control of robot manipulators which are complex systems. This paper presents a novel decentralized model-free robust controller for electrically driven robot manipulators. As a novelty, the proposed controller employs a simple Gaussian Radial-Basis-Function Network as an uncertainty estimator. The proposed netw...
متن کاملLearning Document Image Features With SqueezeNet Convolutional Neural Network
The classification of various document images is considered an important step towards building a modern digital library or office automation system. Convolutional Neural Network (CNN) classifiers trained with backpropagation are considered to be the current state of the art model for this task. However, there are two major drawbacks for these classifiers: the huge computational power demand for...
متن کاملArchitecture Optimization Model for the Multilayer Perceptron and Clustering
This paper presents an approach called Architecture Optimization Model for the multilayer Perceptron. This approach permits to optimize the architectures for the multilayer Perceptron. The results obtained by the neural networks are dependent on their parameters. The architecture has a great impact on the convergence of the neural networks. More precisely, the choice of neurons in each hidden l...
متن کاملComparison of Artificial Neural Network and Multiple Regression Analysis for Prediction of Fat Tail Weight of Sheep
A comparative study of artificial neural network (ANN) and multiple regression is made to predict the fat tail weight of Balouchi sheep from birth, weaning and finishing weights. A multilayer feed forward network with back propagation of error learning mechanism was used to predict the sheep body weight. The data (69 records) were randomly divided into two subsets. The first subset is the train...
متن کامل