Presentation flora

Post on 09-Feb-2017

34 views 0 download

Transcript of Presentation flora

PATTERN RECOGNITION -FLORA EUGIN

(1007-14-508-006)Msc Applied Statistics

PATTERN:

A PATTERN IS A SET OF MEASUREMENTS DESCRIBING A PHYSICAL OBJECT.

PATTERN CLASS:

A PATTERN CLASS (OR CATEGORY) IS A SET OF PATTERNS SHARING COMMON ATTRIBUTES.  

PATTERN RECOGNITIONI t is the study of how machines can: observe the environment, learn to distinguish patterns of interest, make sound and reasonable decisions about

the categories of the patterns.

TWO PHASE PROCESS

Training/Learning

Detecting/Classifying

Learning:How can a machine learn the rule from data? • Supervised learning: A teacher provides a category label or cost for each pattern in the training set. Classification • Unsupervised learning: The system forms clusters or natural groupings of the input patterns (based on some similarity criteria). Clustering • Reinforcement learning: No desired category is given but the teacher provides feedback to the system such as the decision is right or wrong. 

REINFORCEMENT LEARNING Objective: Fly the helicopter Need to make a sequence of good decisions to make it fly Similar to training a pet dog

Every time dog does something good you pat him and say ‘good dog’

Every time dog does some thing bad you scold him saying ‘bad dog’

Over time dog will learn to do good things

Pattern Recognition Process

Data acquisition and sensing:– Measurements of physical variables.– Important issues: bandwidth, resolution , etc.

Pre-processing:– Removal of noise in data.– Isolation of patterns of interest from the background. Feature extraction:– Finding a new representation in terms of features. Classification:– Using features and learned models to assign a pattern to a category.

Post-processing:– Evaluation of confidence in decisions.

FEATUREFeatures are properties of an object. For

e.g.:

WHEN FEATURES ARE GRAPHED“Good” Features “Bad” Features

Feature vector:

Usually a single object can be represented using several features, e.g.– x1 = shape (e.g. nr of sides)– x2 = size (e.g. some numeric value)– x3 = color (e.g. rgb values)– ...– xd = some other (numeric) feature.X becomes a feature vector

x is a point in a d-dimensional feature space

The classical model for Pattern recognition:

1. A Feature Extractor: extracts features from raw data (e.g. audio, image, weather data, etc)

2. A Classifier: receives X and assigns it to one of c categories, Class 1, Class 2, ..., Class c (i.e. labels the raw data).

Example: classify graphic objects according to their shape:

CLASSIFICATION

Apples Oranges

CLASSIFICATION

What is this???

Its an apple!!!

You had some training example or ‘training data’

The examples were ‘labeled’

You used those examples to make the kid ‘learn’ the difference between an apple and an orange

Classification (known categories) Clustering (creation of new categories)

CLASSIFICATION VS. CLUSTERING

17

Category “A”

Category “B”

Clustering(Unsupervised Classification)

Classification(Supervised Classification)

Fields in which Pattern recognition is used:

APPLICATIONS

Safety Identifying finger prints

Facebook uses Face recognition (when you tag, it remembers them…!)

OCR & Handwriting recognition

Speech recognition

APPLICATIONS

Geography :

Military affairs : Bioinformatics : Speech recognition : Computer aided diagnosis :

Earthquake analysis Rocks classification

Aviation photography analysis

DNA sequences analysis Human computer interaction ECG, EEG

AREA Examples

Thank you…!