Page 202 - DCAP208_Management Support Systems

Unit 12: Applications of Neural Network

12.2.1 Pattern Recognition Notes

Pattern Recognition can be defined as the act of taking in raw data and taking an action based on
the category of the pattern. Pattern recognition aims to classify data (patterns) based either on a
priori knowledge or on statistical information extracted from the patterns. The patterns to be
classified are usually groups of measurements or observations, denning points in an appropriate
multidimensional space. This is in contrast to pattern matching, where the pattern is rigidly
specified.
Artificial neural nets can be used for pattern recognition (a task in which humans excel machines!)
The first step is to take an image and convert it to a pixel map (a map of squares with different
intensity of gray color). This pixel map is a two dimensional function with 2 input variables:
pixel position and 1 output variable: intensity of the pixel. We will use standard ANN techniques
to train this 2D function.

Any image can be represented as two-dimensional array, where every element of that array
contains color information for one pixel.

Figure 12.1: Image Represented as Two-dimensional Array

Source: http://wwwteor.mi.infn.it/~rojo/teaching/Milano-NNet-Course-Lecture4.pdf
This information is the input patters for the artificial neural network training.

Figure 12.2: Artificial Neural Network

Source: http://wwwteor.mi.infn.it/~rojo/teaching/Milano-NNet-Course-Lecture4.pdf

LOVELY PROFESSIONAL UNIVERSITY 195