COGS 2160 Lecture Notes - Lecture 2: Operant Conditioning Chamber, Reinforcement Learning, Behaviorism
29 Jan 2017
School
Department
Course
Professor
Document Summary
Bermudez lists four historical landmarks in science of cognition. Idea of algorithmic computation (logic: formal analysis of la(cid:374)guage (cid:894)i(cid:374) people"s (cid:373)i(cid:374)d per a set of rules(cid:895) u(cid:374)co(cid:374)scious, subconscious. Psychologists should study observable phenomena, measurable behavior. Internal mental states: ideas, concepts, thoughts, opinions, hopes, feelings, fears, emotions, etc. Instead focus on bodily movements, reactions, observable responses. Classical conditioning: starts with reaction an organism has (like a reflex) stimulus and a response. Reward (or punishment) is used to strengthen (or diminish) a response: responses that are rewarded tend to be repeated, responses that are punished tend not to be, this is so(cid:373)eti(cid:373)es k(cid:374)ow(cid:374) as the (cid:862)law of effect(cid:863) When response r is made to stimulus s, it is rewarded; when r is not made to s, it is not (or it is punished) Result: r will tend to be repeated to s. Rats in a controlled environment with reward and punishment. All learning is the result of conditioning (i. e. training)
