• Keine Ergebnisse gefunden

Probabilistic Models of Cognition: Generative models

N/A
N/A
Protected

Academic year: 2022

Aktie "Probabilistic Models of Cognition: Generative models"

Copied!
78
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

Probabilistic Models of Cognition:

Generative models

(2)

Table of Contents

(3)

Chapter Content

(4)

Generative Model

(5)

Example: Plinko Machine

(6)

Working Model can be used

for simulation captures some

structure of the world in useful way

(7)

Plinko Machine Demo

Simulate outcomes (data) many times, shape emerges

Reason about ‘shape of expected outcomes’ (with probabilistic concepts)

How to formally describe simulations/working models?

(8)

Building Generative Models

(9)

Examples with Flip

(10)

Flip

(11)

Flip Sum

(12)

Flipping Coins Bend

(13)

Flipping Coins Bend

var bend = function(coin) { return function() {

(coin() == 'h') ?

makeCoin(0.7)() : makeCoin(0.1)() }

}

(14)

Flipping Coins Bend

(15)

Flipping Coins Repeat Sum

(16)

Causal Models in Medical Diagnosis

(17)

Advanced Causal Models in Medical Diagnosis

(18)

Probability Concepts and

WebPPL

(19)

Probability

(20)

Probability Distribution

(21)

Distributions in WebPPL

(22)

Distributions in WebPPL

(23)

Constructing marginal distributions: Infer

(24)

Constructing marginal distributions: Infer

(25)

(26)

The Rules of Probability

(27)

Product Rule

(28)

Product Rule

(29)

Product Rule

(30)

Product Rule

(31)

Product Rule

(32)

Product Rule

(33)

Sum Rule

(34)

Sum Rule

(35)

Sum Rule

(36)

Sum Rule and Product Rule

(37)

Sum Rule and Product Rule

(38)

Sum Rule and Product Rule

(39)

Advanced WebPPL

(40)

Stochastic recursion

(41)

Persistent Randomness: mem

(42)

Persistent Randomness: mem

(43)

Persistent Randomness: mem

(44)

Persistent Randomness: mem

(45)

Example: Intuitive physics

(46)

Example: Intuitive physics

(47)

Example: Intuitive physics

(48)

Example: Intuitive physics

(49)

Example: Intuitive physics

(50)

Summary of Chapter Content

(51)

Exercises

(52)

Exercise 1 a)

(53)

Exercise 1 a)

(54)

Exercise 1 a)

(55)

Exercise 1 b)

(56)

Exercise 1 c)

(57)

Exercise 1 c)

(58)

Exercise 1 c)

(59)

Exercise 1 c)

(60)

Exercise 2

Just one execution of flip

(61)

Exercise 2 b)

(62)

Exercise 2 c)

(63)

Exercise 3

(64)

Exercise 3 a)

(65)

Exercise 3 b)

(66)

Exercise 4 a)

(67)

Exercise 4 b)

(68)

Exercise 4 c)

(69)

Exercise 4 c)

(70)

Exercise 5

(71)

Exercise 5 a)

(72)

Exercise 5 a)

(73)

Exercise 6 a)

(74)

Exercise 6 b)

(75)

Exercise 7 a)

(76)

Exercise 7 b)

(77)

Exercise 8 a)

(78)

Exercise 8 b)

Referenzen

ÄHNLICHE DOKUMENTE

Statistical Shape model: A probability distribution of shapes, learned from example data.. Statistical

• Having a cavity does not depend on whether the patient has a toothache or gum problems. • Does not depend on what the

Sampling

• Rigorous theoretical framework, which helps us to navigate in the space But: Finding good solutions for practical image analysis is hard work!.. There is nothing more practical than

Completion of partial shapes using Gaussian process regression (Project deadline).. • Exercise discussions are mandatory, but

Statistical Shape model: A probability distribution of shapes, learned from example data.. Statistical

Evaluation: Before designing your Gaussian process, think about properties of the model, which allow you to decide whether your model is a good model of the femur shape or

While traffic models that are based on moment generating functions are known for a variety of types of flows, we advance theory by providing corresponding models for a number