AI: Rule-Based Systems and Deep Learning Compared
References:
“Are Rule-Based Systems Still Relevant Today?” - blog post excerpt from Rangarajan Krishnamoorthy's website discusses the continued relevance of rule-based systems in artificial intelligence, particularly in contrast to the rise of machine learning
“Balancing Transparency and Accuracy: A Comparative Analysis of Rule-Based and Deep Learning Models in Political Bias Classification” by Manuel Nunez Martinez , Sonja Schmer-Galunder , Zoey Liu , Sangpil Youm , Chathuri Jayaweera , Bonnie J. Dorr University of Florida, FL,USA {manuel.nunez, s.schmergalunder, liu.ying, youms, chathuri. jayawee, bonniejdorr}@ufl.edu
“Deep Learning” - Wikipedia page on deep learning serves as a comprehensive introduction to the field, outlining its fundamental concepts, historical evolution, and diverse applications.
“Hybrid Approach Combining Machine Learning and a Rule-Based Expert System for Text Categorization” – by Julio Villena-Román, Sonia Collada-Pérez, Sara Lana-Serrano and José C. González-Cristóbal
“RULE-BASED EXPERT SYSTEMS AND BEYOND: AN OVERVIEW” by John Kingston (AIAI-TR-25)
“Rule-Based System or Machine/Deep Learning for My Board Game AI?” - Reddit post from the gamedev community captures a game developer's dilemma in choosing an AI approach for a complex board game.The core question revolves around whether to implement a conventional rule-based system or a more advanced machine/deep learning strategy.
“Rule-based system” - Wikipedia article defines a rule-based system in computer science as a system where knowledge is represented as rules, enabling general reasoning to solve problems. The article further explores the construction of production systems, detailing components like the rule base, inference engine with its match-resolve-act cycle, and working memory.
“Rules-based vs. Deep Learning: A Powerful Synergy in Modern AI” - blog post from ClearObject discusses the power of combining two fundamental approaches in artificial intelligence: rules-based systems, which rely on predefined logic and expert knowledge for transparent decision-making, and deep learning, which uses neural networks to learn complex patterns from vast amounts of data for adaptability and high accuracy.
“Self-learning Symbolic AI: A critical appraisal” – this article argues that while Deep Learning currently dominates the field of Artificial Intelligence, it faces significant limitations. The author contends that the earlier approach of Symbolic AI, which focuses on manipulating symbols and structured knowledge, offers viable solutions to these limitations and has been unfairly overlooked due to the hype surrounding Deep Learning. [Hans Peter Willems MIND|CONSTRUCT March 2021]
“The History of AI: From Rules-based Algorithms to Generative Models” [Jeff Johnston, Last updated on November 12, 2024] - blog post from Lantern, provides a concise overview of the evolution of artificial intelligence. It systematically traces AI's development through key stages: the initial rules-based systems, the emergence of machine learning with its various paradigms, the transformative deep learning revolution, and the recent rise of generative AI and large language models.
