scholarly journals Applying artificial neural networks to top-down construction cost estimating of highway projects at the conceptual stage

2015 ◽  
Author(s):  
Brendon Joseph Gardner
Author(s):  
Margaret A. Boden

Artificial neural networks (ANNs) are made up of many interconnected units, each one capable of computing only one thing. ANNs have myriad applications, from playing the stock market and monitoring currency fluctuations to recognizing speech or faces. ANNs are parallel-processing virtual machines implemented on classical computers. They are intriguing partly because they are very different from the virtual machines of symbolic AI. Sequential instructions are replaced by massive parallelism, top-down control by bottom-up processing, and logic by probability. ‘Artificial neural networks’ considers the wider implications of ANNs and discusses parallel distributed processing (PDP), learning in neural networks, back-propagation, deep learning, and hybrid systems.


Sign in / Sign up

Export Citation Format

Share Document