Traditional Feedforward Network: [Input Layer] ──► [Black-Box Hidden Weights] ──► [Uninterpretable Output] Fu's Hybrid KBCNN Architecture: [Expert Logic Rules] ──► [Initialized Network Architecture] ──► [Refined Empirical Outputs]

(IEEE Transactions on Knowledge and Data Engineering, 1999) — focuses on rule extraction. Knowledge Discovery Based on Neural Networks (Communications of the ACM, 1999). ACM Digital Library hybrid AI models mentioned in these works? Neural Networks in Computer Intelligence | Guide books

The seminal work you are likely looking for is the book Neural Networks in Computer Intelligence

: Rather than starting with random weights, Fu discusses using existing symbolic rules (like "If-Then" logic) to define the initial architecture and weights of a network, allowing it to start from a place of "intelligence" rather than zero. Adaptive Learning

The text is divided into theoretical foundations and practical applications: Theory and Methods

Techniques where the network identifies hidden structures in data, such as Hebbian learning. C. Network Architectures

Fu explains the Sigmoid Activation Function deeply. Use his explanation to write a simple Python function:

Many researchers and students look for digital versions of this classic text for study purposes. While the book was originally published in print, it is sometimes available through academic repositories, library portals, or archive sites.

: One of the book's most unique contributions is its focus on integrating expert rule-based systems with connectionist models.

This is highlighted in chapters dedicated to and "Rule-Generation from Neural Networks" . The core idea is to embed explicit human knowledge into a neural network to improve its learning efficiency, generalization capability, and interpretability—a concept that is highly relevant to today's focus on explainable AI (XAI).

March 1994. Author: LiMin Fu. LiMin Fu. McGraw-Hill, Inc., United States. ISBN : 0079118178. Published: 01 March 1994. Pages: 460. ACM Digital Library gO1HZSRkk1EC (58016015) | PDF - Scribd

: You can access the 1994 edition (ISBN 9780071133197) to read the content online or borrow it digital format. Core Themes and Pedagogical Approach

While modern AI has evolved significantly since 1994, the concepts in "Neural Networks in Computer Intelligence" are essential for understanding the foundations of deep learning. The hybrid approach described by Fu remains at the cutting edge of AI research, aiming to create more interpretable and robust systems.

How to embed expert knowledge directly into network architectures.

I can tailor search terms and alternative resources based on your academic needs. Share public link

Neural Networks In Computer Intelligence Limin Fu Pdf Link Better -