Cascade neural network algorithm with analytical connection weights determination for modelling operations and energy applications

Abstract

The performance and learning speed of the Cascade Correlation neural network (CasCor) may not be optimal because of redundant hidden units' in the cascade architecture and the tuning of connection weights. This study explores the limitations of CasCor and its variants and proposes a novel constructive neural network (CNN). The basic idea is to compute the input connection weights by generating linearly independent hidden units from the orthogonal linear transformation, and the output connection weights by connecting hidden units in a linear relationship to the output units. The work is unique in that few attempts have been made to analytically determine the connection weights on both sides of the network. Experimental work on real energy application problems such as predicting powerplant electrical energy, predicting seismic hazards to prevent fatal accidents and reducing energy consumption by predicting building occupancy detection shows that analytically calculating the connection weights and generating non-redundant hidden units improves the convergence of the network.The proposed CNN is compared with that of the state-of-the-art machine learning algorithms. The work demonstrates that proposed CNN predicts a wide range of applications better than other methods.

Integration of logistic regression and multilayer perceptron for intelligent single and dual axis solar tracking systems

“…This is probably because of CMLP network contains a direct weighted connection from input to output layers, which allows it to learn higher complex patterns as compared to MLP network. 76,77 These additional direct connections also provide linear connections that CMLP network could represent both linear and nonlinear relationships.…”

Hierarchical integrated machine learning model for predicting flight departure delays and duration in series

“…Extensive hyperparameter initializations in machine learning methods may require a lot of trial-and-error experimental work to determine the best optimal structure that has the capability of maximum error reduction. Excessive hyperparameter adjustment along with high dimensional data holding redundant information may cause existing algorithms to converge to a suboptimal solution (Wang et al, 2020). In this study, we propose CNN, named as hyperparameter-free cascade principal component least squares neural network (hyp-free CPCLS), having the characteristics of analytically determining the number of hidden units in hidden layers with no iterative tunning of connection weights.…”