IT (lectures 6& 7) Flashcards
A …… is an abstraction, represented by a set of measurements describing a physical object
Pattern
……. is a set of patterns sharing common attributes
Pattern class
…. refers to some form of adaptation of the classification algorithm to achieve a better response
Learning
…… used to reduce the classification error on a set of training data
Training/ learning
If the classes to which the objects belong are known, the process is called ……..
Supervised learning
If the classes to which the objects belong are unknown, the process is called ……
Unsupervised learning
In ….., no labeled training sets are provided
Unsupervised learning
……. is measurements of physical variables
Data acquisition and sensing
Removal of noise in data is in ….. stage
Pre-processing
Isolation of patterns of interest from the background is in …… stage
Pre-processing
…… partitions an image into regions that are meaningful for a particular task
Segmentation
….. methods, in which similarities are detected
Regio-based
….. methods, in which discontinuities are detected and linked to form continuous boundaries around regions
Boundary-based
Finding a new representation in terms of features is in ….. stage
Feature extraction
….. are characteristic properties of the objects whose value should be similar for objects in a particular class, and different from the values for objects in another class
Features
Features should be ….., they should normally be invariant to translation, orientation (rotation), scale, and illumination
Robust
Features should be ……, the range of values for objects in different classes should be different and preferably be well separated and non-overlapping
Discriminating
Features should be ….., all objects of the same class should have similar values
Reliable
Features should be ….., uncorrelated; as a counter-example, length and area are correlated and it would be wasteful to consider both as separate features
Independent
….. extracts features from raw data
Feature extractor
….. is choosing the most informative subsetbof features, and removing as many irrelevant and redundant features as possible
Feature selection
…… is using features and learned models to assign a pattern to a category
Classification
……. assigns objects to certain categories (or classes) based on the feature information
Classification
Evaluation of confidence in decisions is in ….. stage
Post-processing