Machine Learning, AI, and Soft Computing

Question

What is the difference between these three heavily related fields? Is there one specific whole they are all a part of (aside from CS)?

Answer 1

Machine Learning could be considered a part of AI, however I would classify Machine Learning as the study of creation of semantic models and adaptive behavior with AI being the overall science of systems that intelligent-seeming behavior.

Most of what goes as "AI" is rather simplistic, but highly effective, such as heuristics and the like.

Soft Computing doesn't fell like it has many ML and AI components as it is more about analysis of complex systems. I could be wrong though. As with most things in computer science, the deeper you dig, the more you discover that it's all related.

Answer 2

AI is the intellectual project of trying to capture all aspects of human intelligence in computers. A different project, also called AI, seeks to use human-inspired algorithms to approximate conventionally intractable problems. AI could be said to encompass such fields as robotics, planning, reasoning, learning, and natural language understanding.

Machine learning is the discipline which attempts to improve on a machine's performance of a task, given examples. It could be considered to be within AI's range of interests, but researchers in machine learning need have no intellectual stakes in AI's overall success. Machine learning has a close overlap with statistical physics and certain signal processing topics, and certain formulations strongly overlap planning, control theory, and dynamic programming.

Soft computing involves processes that involve indirect, approximate solutions instead of binary algorithms, widely considered to include such technologies as fuzzy logic, neural networks, and genetic algorithms. There is a broad overlap between these techniques and a certain planning and learning subset of AI, control theory, complex systems theory, etc.

Answer 3

Define machine learning related to soft computing.