Review Sheet_Study_Guide

# Review Sheet_Study_Guide - Computer Science GOFAI: Good Old...

This preview shows pages 1–4. Sign up to view the full content.

ticularly used to rapidly come to a solution that is reasonably close to the best possible answer, or “optimal solution”. Computer Science GOFAI: Good Old Fashion Artificial Intelligence. PSS: (Physical Symbol Systems) Hypothesis 3 parts. A set of symbols, elements which can be manipulated via rules. A set of operations, rules to allow one to manipulate the symbols syntactically. Create, destroy, combine, separate, copy. A body of real world knowledge to supply meaning to symbols.”A physical symbol system (of sufficient size) has the necessary and sufficient means for general intelligent action Logic Theorist (LT): Purley syntactic approach to AI Given a set of symbols, the rules of first order logic, and basic axioms of . Equality. Transitivity. Specification. Able to generate proofs for statements of number theory. Well defined problems (Solve problems with heuristics) Know the goal clearly Know the current situation clearly Know what options are available Problem space for the problem Goal state: What we want to achieve. Initial State: Where we are now Intermediate states: Other possible ways the world could look. Operations: Things I can do in my current state to get either the goal state or some intermediate state. Search for shortest way to goal state

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Brutal force solution: an brute force algorithm for searching a problem space will always find the optimal solution if it exists (takes time and memory to do it). Hill climbing Look one move ahead instead of creating the entire problem space. Compare possibilities for the goal and pick one that gets you the closest. (This is heuristic, it is fast has a low memory load and is usually right. “Eliza” Weizenbaum (1967) Early natural language processing (NLP) system Acted as psychologist, it responded to input from a human at a computer terminal. Initial goal was to show that what looks like language comprehension doesn’t imply any real cognition. (Relatively Stupid)- Function via simple syntactic rules and tables of responses (no semantics at all) SRDLU (Winograd 1971) Responded to uscri input to move geometric solids on a surface. Language processing via syntax only (symbol manipulation) SNePS/CASSIE Statements in natural language can create a mental construct of what the relationship between objects is. Language processing via syntax and some semantics Can learn a large relational database about objects and properties. (Penguins are bird, layeggs) Can do casual reasoning based on those properties and new information. Learn words from surrounding context. Problems for PSS Searle says this is what happens with AI
The program is simply doing syntax Real intelligence/cognition/ mind requires semantics as well Connectionism- Representations of information are encoded as distributed patterns across a large set of individuals “dumb components (nodes/unit) Rules and procedures are encoded as links between nodes in different layers of the system. Learning is achieved by changing the strength of the links between nodes

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## Review Sheet_Study_Guide - Computer Science GOFAI: Good Old...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online