Structural Feature Based Anomaly Detection for Packed Executable Identification - Computational Intelligence in Security for Information Systems - page 285

An Innovative Linkage Learning Based on Differences
in Local Optimums
Hamid Parvin, Behrouz Minaei-Bidgoli, and B. Hoda Helmi
School of Computer Engineering, Iran University of Science and Technology (IUST),
Tehran, Iran
{parvin,b_minaei,helmi}@iust.ac.ir
Abstract. Genetic Algorithms (GAs) are categorized as search heuristics and
have been broadly applied to optimization problems. These algorithms have
been used for solving problems in many applications, but it has been shown that
simple GA is not able to effectively solve complex real world problems. For
proper solving of such problems, knowing the relationships between decision
variables which is referred to as linkage learning is necessary. In this paper a
linkage learning approach is proposed that utilizes the special features of the
decomposable problems to solve them. The proposed approach is called Local
Optimums based Linkage Learner (LOLL). The LOLL algorithm is capable of
identifying the groups of variables which are related to each other (known as
linkage groups), no matter if these groups are overlapped or different in size.
The proposed algorithm, unlike other linkage learning techniques, is not done
along with optimization algorithm; but it is done in a whole separated phase
from optimization search. After finding linkage group information by LOLL, an
optimization search can use this information to solve the problem. LOLL is
tested on some benchmarked decomposable functions. The results show that the
algorithm is an efficient alternative to other linkage learning techniques.
Keywords: Linkage Learning; Optimization Problems, Decomposable Functions.
1 Introduction
GAs are the most popular algorithms in the category of Evolutionary Algorithms
(EAs). These algorithms are widely used to solve real-world problems. However
when it comes to solve difficult problems, GA has deficiencies. One of the main
problems of simple GAs is their blindness and oblivion about the linkage between the
problem variables. It is long time that the importance of the linkage learning is
recognized in success of the optimization search. There are a lot of linkage learning
techniques. Some are based on perturbation methodology, some are categorized in the
class of probabilistic model building approaches and some are the techniques that
adapt the linkages along with the evolutionary process by employing special operators
or representations.
In this paper a new linkage learning approach, which is called LOLL is proposed.
The proposed algorithm as its title implies, does not fall in the above mentioned
categories, but it is a linkage group identification approach which tries to identify