Neural Networks in Computer Intelligence by LiMin Fu is a seminal 1994 text that explores the integration of connectionist models (neural networks) with traditional artificial intelligence. You can access digitized versions of the book through the Internet Archive Bridging the Gap: Neural Networks Meets Symbolic AI
Training neural networks involves adjusting the model's parameters to minimize a loss function. Common training algorithms include:
Use this book to understand "shallow" networks. Once you understand Backpropagation as explained by Fu, compare it to modern Deep Learning textbooks. You will realize that the core logic has not changed, only the scale (layers) and the computing power.
: Emphasis on integrating knowledge-based systems with connectionist models.
