A Sequential Learning Algorithm for Complex-Valued Self-Regulating Resource Allocation Network-CSRAN

Abstract

This paper presents a sequential learning algorithm for a complex-valued resource allocation network with a self-regulating scheme, referred to as complex-valued self-regulating resource allocation network (CSRAN). The self-regulating scheme in CSRAN decides what to learn, when to learn, and how to learn based on the information present in the training samples. CSRAN is a complex-valued radial basis function network with a sech activation function in the hidden layer. The network parameters are updated using a complex-valued extended Kalman filter algorithm. CSRAN starts with no hidden neuron and builds up an appropriate number of hidden neurons, resulting in a compact structure. Performance of the CSRAN is evaluated using a synthetic complex-valued function approximation problem, two real-world applications consisting of a complex quadrature amplitude modulation channel equalization, and an adaptive beam-forming problem. Since complex-valued neural networks are good decision makers, the decision-making ability of the CSRAN is compared with other complex-valued classifiers and the best performing real-valued classifier using two benchmark unbalanced classification problems from UCI machine learning repository. The approximation and classification results show that the CSRAN outperforms other existing complex-valued learning algorithms available in the literature.

Adaptive division of labor particle swarm optimization

“…In human cognitive psychology (Nelson & Narens, 1990), self-regulated learning controls the learning process through proper planning and selection of appropriate learning strategies; therefore, regulating the actions of a learner to achieve the desired results is an effective strategy. The machine learning literature (Suresh, Dong, & Kim, 2010;Suresh, Savitha, & Sundararajan, 2011) also reveals that the learning algorithm that employs self-regulation exhibits better generalization performance over other algorithms. Given these facts, ADOLPSO can be employed as an effective learning algorithm to train most machine learning techniques, including artificial neural network (G. Das, Pattnaik, & Padhy, 2014), neuro-fuzzy network (Chatterjee & Siarry, 2007), and fuzzy rule-based system (García-Galán, Prado, & Muñoz Expósito, 2015), which are the most essential blocks to establish modern intelligent and expert systems.…”

Scaled Conjugate Gradient Learning for Complex-Valued Neural Networks

“…Fully complex functions treat complex numbers as a whole, and not the real and imaginary parts separately, as the split complex functions do. This problem was used as a benchmark to test the performance of different complex-valued neural network architectures and learning algorithms, for example in [20][21][22]25].…”

“…FC-RBF [20,21] 3.61e-6 9.00e-6 FC-RBF with KMC [21] 2.01e-6 1.87e-6 Mc-FCRBF [22] 2.50e-5 2.56e-6 CSRAN [25] 9.00e-6 9.00e-6 CMRAN [20,21] 4.60e-3 4.90e-3…”

Short-Term Load Forecasting Using Adaptive Annealing Learning Algorithm Based Reinforcement Neural Network

“…Besides, researchers have proposed sequential learning algorithms for resource allocation networks to enhance the convergence of the training error and computational efficiency [25]. A reinforcement learning method based on adaptive simulated annealing has been adopted to improve a decision making test problem [26].…”