ML & Knowledge-Based Systems
ML & Knowledge-Based Systems
Short description explaining what the learner will understand after completing this study map.
Two Ways to Build Intelligent Systems
Machine learning systems learn patterns from data. Knowledge-based systems instead rely on explicitly represented knowledge, often combined with rules or an inference mechanism to reach conclusions.
Core idea: machine learning derives behavior from data and experience, while a knowledge-based system works with knowledge that has been explicitly represented for the system to use.
Essential Terms
Tap each card to compare the building blocks of machine learning and knowledge-based systems.
How the Two Approaches Work
The two approaches obtain their capability differently. Machine learning identifies patterns from examples or data, whereas a knowledge-based system reasons using stored knowledge and explicit relationships or rules.
Learned Patterns vs Explicit Knowledge
The key contrast is where the system's useful information comes from. In machine learning, patterns are learned from data. In a knowledge-based approach, relevant knowledge is represented explicitly.
Machine Learning
Uses data and algorithms to learn patterns that support predictions, classifications, recommendations, or other outputs.
Knowledge-Based System
Uses explicitly represented domain knowledge. Rules or an inference mechanism can apply that knowledge to facts in order to reach conclusions.
From Input to Conclusion
Both approaches can transform inputs into useful outputs, but the internal route differs: one relies on learned patterns, while the other applies explicit knowledge and reasoning.
KBS: FACTS + KNOWLEDGE/RULES → INFERENCE → CONCLUSION
Why Combine the Approaches?
Machine learning can discover useful patterns that are difficult to specify manually. Knowledge-based reasoning can make explicit domain knowledge available to the system. In some solutions, combining learned patterns with explicit knowledge can provide complementary capabilities.
System Explorer
Explore the difference between learning from data and reasoning with explicitly represented knowledge.
Key Takeaways
Focus on the source of intelligence: learned patterns from data versus explicit knowledge and reasoning.
Core pattern: machine learning learns patterns from data; knowledge-based systems use explicitly represented knowledge and reasoning to reach conclusions.
Common mistake: a knowledge-based system is not simply a database. Stored knowledge becomes useful when the system can apply or reason with it to solve a problem.
Knowledge Check
These questions test whether you can distinguish machine learning from knowledge-based reasoning and recognize how the approaches can complement one another.