Abstract. human behavior identification. Numerous applications of our daily

Abstract. This chapter introduces a hybrid algorithm for classifying
both general human behavioral patterns as well as specific actions. Behavior
modeling is a significant research area in this field and has been
an active research topic for several applications. Achieving a high accuracy
classification model has also been a research target which derived
many efforts for improving classification performance. Tuning classifiction
parameters and data features are vital methods for obtaining accurate
classifier in any application. This work focuses on tuning classification
parameters of SVM classifier and optimizing features set using
Elephant Herding Optimization Algorithm (EHO). The proposed classification
model will be validated using behavioral data obtained from a
motion capturing camera system known as VICON to accurately classify
normal and aggressive behavioral patterns and specific actions.
Keywords: Behavior, Classification, Optimization, Vicon, Elephant Herding (EHO)
, Support Vector Machines (SVMs).
Optimizing classification parameters and classification features is a rich research
topic. In this chapter meta-heuristic swarm-based algorithm is devoted to finetune
classification parameters to get best performance for human behavior identification.
Numerous applications of our daily life involve machine learning procedures.
Currently, there are many sources that are rich of behavioral information
sich as images, video records, and many types of sensors where the internet of
things comes (IOT) 1,2,3. Such data is analyzed and introduced to aid real life
applications such as security, entertainment and medical diagnosis4. Lately,
surveillance systems are surrounding us everywhere in the street, markets, and
organizations from simplest to most complicated ones 5,6. Modern systems are
supported with detection techniques for dubious behavioral patterns, classifying
people motions into normal and abnormal patterns to respond properly and in
real time 7.
Many research efforts have been applying machine learning techniques (MLs)
for the purpose of human actions identification. Support Vector Machines (SVMs),
Neural networks and Support Vector Regression (SVR) are examples 8. The
proper adjustment of learning parameters is a problem facing model designers
who sake for acheiving a model that accomplishes high generalization as well as
high classification accurac 9. Other main problem faceing ML model designers
is feature selection. Feature selection is concerned with filtering the input set of
features to only most discriminant and most features relevant to the data. In this
chapter, we discuss how Swarm Intelligence (SI) had a great share of machine
learning optimization research in past years 10,11. In this work we show how
SI could be employed to solve the problem of classification parameters tuning as
well as features selection using selected SI algorithms.