Perception: —)Memory¹  and Eudoxa²——) Statements—) Universals —-) Logic and Science³         Particulars—-) Practical Wisdom⁴

1. Memory allows learning and sociality. No memory, no connections. Repeated memories lead to experience and eudoxa.
2. Eudoxa: Good opinions based on collective human experience: the social basis of learning, a review of what is commonly thought to be correct, as the basis of improving the recognition of true (developed over time and experience) principles and testing of those principles. Also places principles in context. Note: learning is more than pattern recognition, it requires context. For Aristotle, the key to context is experience, organized as the four causes: formal, material, efficient and final (Meta. Bk. one, chapter three).
3. Logic is the structure of science (episteme); it is the organization of what can be demonstrated. For Aristotle logic is primarily demonstration through syllogisms that organize universal statements:
   1. E.g., Barbara: the name of the structure of a universal affirmative syllogism:
      1. All animals have a metabolism.
      1. All humans are animals.
      1. Therefore, all humans have a metabolism.
4.  Practical wisdom (phronesis): this is experientially based knowledge of activities and things that hold for the most part. Because of change, there can be no universal demonstration of phronesis, but its generalizations allow successful or unsuccessful efforts to accomplish plans or maintain relationships. This type of knowledge ranges from the practical implications of the scientific knowledge of human nature (e.g., the art of medicine) to emotional intelligence to the basis of friendship to politics.

For more: google Aristotle’s epistemology and I recommend The Stanford Encyclopedia of Philosophy and the Encyclopedia Britannica. For Aristotle’s own writings (and no one says it better or more clearly), start with the Metaphysics Book One, chapters one and two, and the first few lines of chapter 3; and Book One of the Nicomachean Ethics.

Note:  From an Aristotelian perspective, there are (at least) two serious problems with artificial intelligence: the struggle to get AI to recognize the context for principles and generalizations, and the struggle to adjust to change. In short, pattern recognition generates static results on the basis of surface identification. By contrast Aristotelian knowledge in embedded in the experience humans have, which is only partly made up of pattern recognition, and at this point experience and its fruits remain more sensitive and subtle than pattern recognition no matter how expansive the AI data set is. AI is very perceptive of static detail, making it good at radiology but not so good at treating patients.