Lect2 - Today: Models and beginning search Chapters 3 &...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
Today: Models and beginning search How to read the text Skim the assignment Read interesting/important parts Listen to lectures; ask questions Read (again?) the interesting / important parts Ask questions at the beginning of class and at office hours 1 Lecture 2
Background image of page 1

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

View Full DocumentRight Arrow Icon
Unified view of AI as Model-based Reasoning What is reasoning? Making a decision Drawing a conclusion Choosing an action Developing an interpretation What is a “model?” A stand-in for the real thing Must be mathematically precise: quantifiable computational properties NB in logic there is a different technical usage 2 Lecture 2
Background image of page 2
Inference with Computational Models Inference: explicit new justified beliefs Model design Driven by task As simple as possible Classification From input features Infer a class label Problem Solving Decisions/actions (often a sequence) Achieve a goal In a moment, the search example of cryptarithmetic 3 Lecture 2
Background image of page 3

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

View Full DocumentRight Arrow Icon
Designing a Model Distinctions to notice Real world is infinitely subtle and complex Ideally notice all and only necessary distinctions (leave out almost everything) States / attributes / features Dynamics change Operators / successor states / transition function (R When: preconditions How: effects Often change = world change but not always Inference change in explicit beliefs about world True? Useful? 4 Lecture 2
Background image of page 4
AI Models TODAY’s AI Models are often empirically driven and statistically sophisticated Machine Learning plays an important role Weak prior knowledge / analytic commitments Relies on empirical evidence Purely Analytic (deep, brittle) Primarily Empirical (robust, weak) Dominant Today 5 Lecture 2
Background image of page 5

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

View Full DocumentRight Arrow Icon
Analytic / Emprical Models can be purely analytic Models cannot be purely empirical For us “analytic” means purely analytic; “empirical” means not purely analytic Empirical models always have some analytic component; they use some observations of the world 6 Lecture 2
Background image of page 6
Why Use Empirical? Necessity:
Background image of page 7

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

View Full DocumentRight Arrow Icon
Image of page 8
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 38

Lect2 - Today: Models and beginning search Chapters 3 &...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online