1. Introduction to

unknownlippsΤεχνίτη Νοημοσύνη και Ρομποτική

16 Οκτ 2013 (πριν από 4 χρόνια και 22 μέρες)

59 εμφανίσεις

1. Introduction to

Pattern Recognition and

Machine Learning.

Prof. A.L. Yuille.

Dept. Statistics. UCLA.

Stat 231. Fall 2004.


Examples of Patterns.

Discriminate/Decisions about Patterns.

Schools of Pattern Recognition.

Learning Theory.

What are Patterns?

Laws of Physics & Chemistry generate patterns.

Patterns in Astronomy.

Humans tend to see patterns everywhere.

Patterns in Biology.

Applications: Biometrics, Computational
Anatomy, Brain Mapping.

Patterns of Brain Activity.

Relations between brain activity, emotion,
cognition, and behaviour.

Variations of Patterns.

Patterns vary with

Speech Patterns.

Acoustic signals.

Goal of Pattern Recognition.

Recognize Patterns. Make decisions about

Visual Example

is this person happy or

Speech Example

did the speaker say
“Yes” or “No”?

Physics Example

is this an atom or a

Applications of Pattern Recognition.

Handwritten digit/letter recognition

Biometrics: voice, iris, fingerprint, face, and gait recognition

Speech recognition

Smell recognition (e
nose, sensor networks)

Defect detection in chip manufacturing

Interpreting DNA sequences

Fruit/vegetable recognition

Medical diagnosis

Terrorist Detection

Credit Fraud Detection

Credit Applications.

… …

Two Extreme Approaches

Generative Methods:

Determine models of how patterns are formed.

Use these models to perform discrimination.

Pattern Theory. Grenander.

Discriminative Methods

Don’t model pattern formation.

Instead extract features from patterns and make
decision using these features.

Example: Salmon versus Sea Bass.

Generative methods
attempt to model the
full appearance of
Salmon and Sea

methods extract
features sufficient to
make the decision
(e.g. length and

Fish Features. Length.

Salmon are usually shorter than Sea Bass.

Fish Features. Lightness.

Sea Bass are usually brighter than Salmon.

Decision Boundaries.

Classify fish as
Salmon or Sea Bass
based on a decision
boundary in feature

Generative Models for Speech.

Stochastic Grammars for Speech & Natural
Language. (Manning & Schutze).

Bayes Decision Theory

Bayes Decision Theory

gives a framework for
Generative and Discriminative approaches.

Current Wisdom:

(i) Discriminative methods are simpler,
computationally faster, and easier to apply.

(ii) Generative methods are needed for most
complex problems.

Hybrid methods are increasingly popular.

Stat 231 concentrates on Discriminative Methods
and simple Generative Models.

Other courses by Prof.s Zhu & Yuille deal with
complex Generative Models.

Learning Theory.

Both Generative and Discriminative
methods require training data to learn the
models/features/decision rules.

Machine Learning concentrates on
learning discrimination rules.

Key Issue: do we have enough training
data to learn?

Course Elements.

Bayes Decision Theory as theoretical

Simple discriminative and generative

Machine Learning.

Advanced Discriminative Methods.