IRJET- Pattern Recognition Process, Methods and Applications in Artificial Intelligence

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 11 | Nov 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1162
Pattern Recognition Process, Methods and Applications in Artificial
Intelligence
Susmita Karyakarte1, Prof. Ila Savant2
1Student,Department of Computer Engineering, Marathwada Mitra Mandal’s College of Engineering, Pune,
Maharashtra, India
2Professor, Department of Computer Engineering, Marathwada Mitra Mandal’s College of Engineering, Pune,
Maharashtra, India
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - Pattern recognition can be seen as a classification
process.Pattern recognition has become more and more
popular and important to us attractingattentionasitfinds it’s
applications into widespread areas of research. Pattern
recognition plays an important role: reading texts, identifying
people, retrieving objects, or finding the way in a city. Once
patterns are established, however learned, we are able to
classify new objects or phenomena into a class of known
patterns. In Artificial Intelligence using Pattern Recognition
we can explicitly program machines to learn these these
patterns by themselves and use this information to perform
specific tasks. It’s the very first step involved in developmentof
most AI systems that makes it an inseparable part of the
process thus finding its way into varioussubdomainsofAI. The
objective of this paper is to discuss various steps involved in
pattern recognition and how theyarecommontoalmostevery
system application of artificial intelligence and which
algorithms are popularly used to implement pattern
recognition in various subdomains.
Key Words: Pattern Recognition, Artificial Intelligence,
Preprocessing, FeatureExtraction,Classification, Neural
Networks, Support Vector Machine, Image Processing,
Data Analytics.
1. INTRODUCTION
The field of pattern recognition is concerned with the
automatic discovery of regularitiesindata throughtheuseof
computer algorithms and with the use of these regularities
to take actions such as classifying the data into different
categories. Its ultimate goal is to optimally extract patterns
based on certain conditions and to separate one class from
the others. The application of Pattern Recognition can be
found everywhere. Examples includediseasecategorization,
prediction of survival rates for patients of specific disease,
fingerprint verification, face recognition, irisdiscrimination,
chromosome shape discrimination, optical character
recognition, texture discrimination, speech recognition,etc.
The design of a pattern recognition system should consider
the application domain. It is striking and interesting to
observe that artificial recognition devices, especially the
ones that learn from examples, are almost not, or just
superficially based on a modeling of the human perception
and learning abilities. One of the reasons is that the artificial
systems may serve different purposes and they need to be
more stable and should sometimes be faster and larger, at
the cost of a reduced flexibility.
Human beings are pattern recognizers, not just because of
this recognition ability, but especially because we are aware
of it. We can handle it and also teach the patterns to others
and discuss with them our observations. The ability to judge
the similarity between objects or events is called
generalization. The question of how our mind travels from
observations to memory and to generalizationistherebythe
basic scientific question of pattern recognition and how this
process can be integrated in and taught to a computer, a
challenge for it’s algorithms. Pattern Recognition can be
implemented with the use of Machine Learning algorithms.
These algorithms perform classification of data based on
knowledge already gained or on statistical information
extracted from patternsand/ortheirrepresentation.Pattern
recognition is the ability to detect arrangements of
characteristics or data that yield information about a given
system or data set. Predictive analytics in data science work
can make use of pattern recognition algorithms to isolate
statistically probable movements of time seriesdata intothe
future. In a technological context, a pattern might be
recurring sequences of data over time that can be used to
predict trends, particular configurations of features in
images that identify objects,frequentcombinationsof words
and phrases for natural language processing (NLP), or
particular clusters of behaviour on a network that could
indicate an attack — among almost endless other
possibilities. A basic PR systemcompletelyreliesondata and
derives any outcome or model from data all by itself
recognizing familiar patterns quickly and accurately. The
basic components of a pattern recognition process are
preprocessing, feature extraction, and classification used in
every PR system which are discussed further in detail.
2. Pattern Recognition Process
Pattern recognition has been under constant development
for many years. It includes lots of methods impelling the
development of numerous applications in different fields.
The basic components in pattern recognition are
preprocessing, feature extraction, and classification. Once
the dataset is acquired, it is preprocessed, so that itbecomes
suitable for subsequent sub-processes. The next step is
feature extraction, in which, the dataset is converted into a
set of feature vectors which are supposed to be

representative of the original data. These features are used
in the classification step to separate the data points into
different classes based on the problem.[1]
Figure 1: Basic pattern recognition process common to
most models
Figure 2: Pattern recognition process in a Statistical PR
Model
2.1 Preprocessing
The role of preprocessing is to segment the interesting
pattern from the background. It is used to reduce variations
and produce a more consistent set of data. Preprocessing
should include some noise filtering, smoothing and
normalization to correct the image from different errors,
such as strong variations in lighting direction and intensity.
In some applications,segmentationoftheinterestingpattern
of a given image from the background is very important, for
example, dealing with diseases detection in agriculture
applications needs segmentationoftheinfectedregionof the
diseased plant images.
2.2 Feature Extraction
Feature extraction is used to overcome the problem of high
dimensionality of the input set in pattern recognition.
Therefore, the input data will be transformed into a reduced
representation set of features, also named feature vector.
Only the relevant information from the input data should be
extracted in order to perform the desired task using this
reduced representation instead of the full size input.
Features extracted should be easily computed, robust,
rotationally invariant, and insensitive to various distortions
and variations in the images. Then optimal features subset
that can achieve the highest accuracy results should be
selected from the input space.
Two kinds of features are used in pattern recognition
problems. One kind of features has clear physical meaning,
such as geometric or structural and statistical features.
Another kind of features has no physical meaning. We call
these features mapping features. The advantage of physical
features is that they need not deal with irrelevant features.
The advantage of the mapping features is that they make
classification easier because clear boundaries will be
obtained between classes but increasing the computational
complexity. Most of feature selection algorithms involve a
combinatorial search through the whole space. Usually,
heuristic methods, such as hill climbing, have to be adopted,
because the size of input space is exponential in the number
of features. Other methods divide the feature space into
several subspaces which can be searched easily. There are
basically two types of feature selection methods filter and
wrapper. Filters methods select the best features according
to some prior knowledge without thinking about the bias of
further induction algorithm. So these methods performed
independently of the classification algorithm or its error
criteria.
In feature extraction, most methods are supervised. These
approaches need some prior knowledgeand labeledtraining
samples. There are two kinds of supervised methods used:
Linear feature extraction and nonlinear feature extraction.
Linear feature extraction techniques include Principal
Component Analysis (PCA), Linear Discriminant Analysis
(LDA), projection pursuit, and Independent Component
Analysis (ICA). Nonlinearfeature extractionmethodsinclude
kernel PCA, PCA network, nonlinear PCA, nonlinear auto-
associative network, Multi-Dimensional Scaling (MDS) and
Self-Organizing Map (SOM), and so forth.
2.3 Classification
During the classification task, the system uses the features
extracted in the previous stage from each of the patterns to
recognize them and to associate each one to its appropriate
class. Two types of learning procedure are found in the
literature. The classifiers that contain the knowledgeofeach
pattern category and also the criterion or metric to
discriminate among patterns classes, which belong to the
supervised learning. The unsupervisedlearningin whichthe
system parameters are adapted using only the information
of the input, and constrained byprespecifiedinternal rules,it
attempts to find inherent patterns in the data that can then
be used to determine the correct output value for new data
instances.
For example, when determining whether a given image
contains a face or not, the problem will be a face/non-face
classification problem. Classes, or categories, are groups of
patterns having feature values similar according to a given
metric. Pattern recognition is generally categorized
according to the type of learning used to generatetheoutput
value in this step. This step enables us to recognize an object
or a pattern by using some characteristics(features)derived
from the previous steps. It is the step which attempts to
assign each input value of the feature vectortooneofa given
set of classes. There are differentclassificationmethodsused
in the pattern recognition process that haveit'sownabilities
and characteristics. They are listed and describedinshort as
follows-
o Fuzzy ART
o Neural Networks

o Markov Random Fields
o Support Vector Machine
2.3.1 Fuzzy ART
Fuzzy ART neural networks can be used as an unsupervised
vector classifier. Adaptive Resonance Theory (ART) is
compatible with the human brain in processinginformation,
it has the ability to learn and memorize a large number of
new concepts in a manner that does not necessarily cause
existing ones to be forgotten. ART is able to classify input
vectors which resemble each other according to the stored
patterns. Also, it can adaptively create a new corresponding
to an input pattern, if it is not similar to any existing
category. ART1 was the first model ofART,itcanstablylearn
how to categorize binary input patterns presented in an
arbitrary order. Plus, Fuzzy sets theory can imitate thinking
process of human being widely and deeply. [6]So the Fuzzy
ART model, which incorporates computationsfromfuzzy set
theory into the ART1 neural network, is capable of rapid
stable learning of recognition categories in response to
arbitrary sequences of analog or binary input patterns.
2.3.2 Neural Networks
The neural approach appliesbiological conceptsto machines
to recognize patterns. It is a promising and powerful tool for
achieving high performance in pattern recognition. The
outcome of this effort is the invention of artificial neural
networks which is set up by the elicitation of the physiology
knowledge of human brain. Neural networks are composed
of a series of different, associate unit. It is about mapping
device between an input set and an output set. Since the
classification problem is a mappingfromthe featurespaceto
some set of output classes, we can formalize the neural
network, especially two-layer Neural Network asa classifier.
A basic Artificial Neural Network(ANN) can consist of the
following layers-the input layer, one or more hidden layers
and the output layer.[7]While the usual scheme choosesone
best network from amongst the set of candidate networks,
better approach can be done by keeping multiple networks
and running them all with an appropriate collectivedecision
strategy. Multiple neural networks can be combined for
higher recognition rate.
Figure 3: A simple Artificial Neural Network(ANN)
2.3.3 Markov Random Fields
Markov random fields (MRFs) are a kind of probabilistic
model which encodes the model structure as an undirected
graph. Two variables are connected by an edge if they
directly influence each other. Markov random field (MRF)
models are multi-dimensional in nature, for pattern
recognition, they combine statistical and structural
information. States are used to model the statistical
information, and the relationships between states are used
to represent the structural information. Only the best set of
states should be considered. The global likelihood energy
function can be rewritten with two parts, one used to model
structural information that is described by the relationships
among states, and the second models the statistical
information because it is an output probability for the given
observation and state.[9] The recognition process is to
minimize the likelihood energy function that is the
summation of the clique functions. It consists of an
undirected graph G = (N, E) in which the nodes N represent
random variables and the edges E encode conditional
independence relationships via some defined rules.
2.3.4 Support Vector Machine (SVM)
The Support Vector Machine (SVM) classifier has been
proved to be very successful in many applications.
The strength of the SVM is its capacity to handle not only
linearly separable data, but also non-linearly separable data
using kernel functions. The kernel function can map the
training examples in input space into a feature space such
that the mapped training examples are linearly separable.
The frequently used SVM kernels are: polynomial, Gaussian
radial basis function, exponential radial basis function,
spline, wavelet and autocorrelation wavelet kernel.
Theoretically, features with any dimension can be fed into
SVM for training, but practically, features with large
dimension have computation and memory that cost to the
SVM training and classification process, therefore, feature
extraction and selection is a crucial step before the SVM
classification.[4][5]
3. Pattern Recognition Models and their Applications
in Artificial Intelligence
General purpose pattern recognition is a very difficult
problem hence use of object models, constraintsandcontext
is necessary for identifying complex patterns and varies
depending upon the application. No single recognition
approach has been found to be optimal for all pattern
recognition problems. Following are the three main models
of Pattern Recognition –
 Statistical Model: It identifies a specificpieceofdata
by studying and learning from the alreadyclassified
training data using statistics. This model
uses supervised machine learning methods.

 Syntactic / Structural Model: Used to define a more
complex relationship between elements (for
example, parts of speech). This model uses semi-
supervised machine learning methods.
 Template Matching Model: Matches the object's
features with the predefined template and identify
the object by proxy. One of the uses of such model is
plagiarism checking.
Artificial Intelligence
Artificial intelligence (AI) is the simulation of human
intelligence processes by machines, especially computer
systems. These processesincludelearning(theacquisitionof
information and rules for using the information), reasoning
(using rules to reach approximate or definite conclusions)
and self-correction. AI is ubiquitous today, used to
recommend what you should buy next online, tounderstand
what you say to virtual assistants such as Amazon's Alexa
and Apple's Siri, to recognise who and what is in a photo, to
spot spam, or detect credit card fraud.AI has a number of
subdomains such as Computer Vision, Deep Learning,
Natural Language Processing(NLP),Data Analytics, etc. In
this paper we will be discussing applications of pattern
recognition in the NLP, Image Processing and Data Analytics
subdomains.
3.1 PR in Natural Language Processing(NLP)
Natural language processing techniques train computers to
understand what a human speaks. Natural language
processing gives machines the ability to read and
understand the languages that humans speak. [8]NLP
research aims to answer the question of how peopleareable
to comprehend the meaning of a spoken/written sentence
and how people understand what happened, when and
where that happened or what isanassumption,belieforfact.
The Syntactic / Structural Model of Pattern Recognition
proves to be the most useful in NLP systems to classify
words and comprehend the given textual data.Thefollowing
diagram represents the flow of a Syntactical model –
Figure 4: Syntactic / Structural Model in PR
Applications of the model in NLP are numerous and can be
described as follows:-
 Text analysis - For content categorization, topic
discovery and modeling (content marketing tools
like Buzzsumo use this technique).
 Text summarization andcontextual extraction – For
finding the meaning of the text. There are many
online tools for this task, for example, Text
Summarizer.
 Text generation - For chatbots and AI Assistants or
automated content generation (for example, auto-
generated emails, Twitterbot updates, etc.);
 Text translation - In addition to text analysis and
word substitution, the engine also uses a
combination of context and sentiment analysis to
make closer matching recreation of the message in
the other language. The most prominent example
is Google Translate.
 Text correction and adaptation - In addition to
correcting grammar and formal mistakes, this
technique can be used for the simplification of the
text - from the structure to the choice of
words. Grammarly, a startupthatprovidessoftware
for text correction is one of the most prominent
examples of such NLP pattern recognition uses.
3.2 PR in Image Processing, Segmentation and Analysis
Pattern recognition is used to give human recognition
intelligence to machine which is required in image
processing. Template matching model is used in digital
image processing for finding small parts of an image which
match a template image.[2] It can be used in manufacturing
control, a way to navigate a mobile robot or as a way to
detect edges in images.[3] The following diagram helps to
understand the concept of template matching :-
Figure 5: Template Matching
Applications are found for the detection of plant diseases to
minimize the loss, and achieve intelligent farming. Plant
diseases are one of the most important reasons that lead to

the destruction of plants and crops. Detectingthosediseases
at early stages enables us to overcome and treat them
appropriately. The development of an automation system
using pattern recognition for classifying diseases of the
infected plants is a growing research area in precision
agriculture. For the identification of the visual symptoms of
plant diseases, texture, color, shape and position of the
infected leaf area might be used as discriminator features.
Diseased regions such as spots, stains or strikes should be
identified, segmented, preprocessed and a set of features
should be extracted from each region.
As applications are found for identification of plantdiseases,
the same way pattern recognition can be used for early
identification of diseases like cancer in patients with likely
symptoms to help in medical diagnosis.
3.3 PR in Data Analytics
Pattern Recognition technology and Data Analytics are
interconnected to the point of confusion between the two.
An excellent example of this issue is stock market pattern
recognition software,whichisactuallyananalytics tool.Data
analytic models are excellent examples of model following
the statistical approach of pattern recognition.Inthecontext
of data analytics, pattern recognition is used to describe
data, show its distinct features (i.e., the patterns itself) and
put it into a broader context.
Let’s look at two prominent use cases:
 Stock market forecasting - Pattern Recognition is
used for comparative analysis of the stock
exchanges and predictions of the possible
outcomes. Yard Charts use this pattern recognition
analysis.
 Audience research - Pattern Recognition refers to
analyzing available user data and segmenting it by
selected features. Google Analytics provides these
features.
4. CONCLUSIONS
In this paper we elucidate pattern recognition in the round,
including the definition, the methods of pattern recognition,
the basic models of pattern recognition and it’s applications
in various subdomains of artificial intelligence. The
applications of pattern recognitionareincreasingdayby day
thanks to the immense progress and development in
machine learning algorithms used in all the PR models. It
aims to classify a pattern into one of a number of classes. It
appears in various fields like natural language processing
(NLP), image processing, data analytics, etc.Inthiscontext,a
challenge consists of finding some suitable description
features since commonly, the patterntobeclassifiedmust be
represented by a set of features characterizing it. These
features must have discriminative properties: efficient
features must be affined insensitive transformations. They
must be robust against noise and against elastic
deformations. The feature extracting method and the
classifier should depend on the application itself. General
purpose pattern recognition is very difficult and no single
recognition approach has been found to be optimal for all
pattern recognition problems. Hence future work will be
done by searching for the right methods todevelopa pattern
recognition system that proves to be optimal for most
pattern recognition problems.
REFERENCES
[1]Elie, Saliba. (2013). An overview of Pattern Recognition.
[2] Girolami, M.; Chao He; “Probability density estimation
from optimally condensed data samples” Pattern Analysis
and Machine Intelligence, IEEE Transactionson, Volume: 25,
Issue: 10, pp: 1253 – 1264,Oct. 2003.
[3] Meijer, B.R.; “Rules and algorithms for the design of
templates for template matching”, Pattern Recognition,
1992. Vol.1. Conference A: Computer Vision and
Applications, 11thIAPRInternational Conferenceon,pp: 760
– 763, Aug.1992.
[4] Vapnik, V., The Nature of Statistical Learning Theory,
Springer, 1995.
[5] Julia Neumann, Christoph Schnorr, “SVM-based feature
selection by direct objective minimization”, 2004.
[6] Mun-Hwa Kim, Dong-Sik Jang, and Young-Kyu Yang. A
robust-invariant pattern recognition model using fuzzy art.
Pattern Recognition, 34(8):1685{1696, 2001.
[7] Vladimir N Vapnik. An overview of statistical learning
theory. IEEE transactions on neural networks,
10(5):988{999, 1999.
[8] L. R. Ruiz, "Interactive Pattern Recognition applied to
Natural Language Processing," Thesis, 2010.
[9] Jinhai Cai and Zhi-Qiang Liu. Pattern recognition using
markov random field models. Pattern Recognition,
35(3):725{733, 2002.

IRJET- Pattern Recognition Process, Methods and Applications in Artificial Intelligence

Recommended

More Related Content

What's hot (19)

Similar to IRJET- Pattern Recognition Process, Methods and Applications in Artificial Intelligence (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Pattern Recognition Process, Methods and Applications in Artificial Intelligence