Papers by Surafel L. Tilahun
Lecture Notes in Computer Science, 2012
Multilevel optimization problems deals with mathematical programming problems whose feasible set ... more Multilevel optimization problems deals with mathematical programming problems whose feasible set is implicitly determined by a sequence of nested optimization problems. These kind of problems are common in different applications where there is a hierarchy of decision makers exists. Solving such problems has been a challenge especially when they are non linear and non convex. In this paper we introduce a new algorithm, inspired by natural adaptation, using (1+1)-evolutionary strategy iteratively. Suppose there are k level optimization problem. First, the leader's level will be solved alone for all the variables under all the constraint set. Then that solution will adapt itself according to the objective function in each level going through all the levels down. When a particular level's optimization problem is solved the solution will be adapted the level's variable while the other variables remain being a fixed parameter. This updating process of the solution continues until a stopping criterion is met. Bilevel and trilevel optimization problems are used to show how the algorithm works. From the simulation result on the two problems, it is shown that it is promising to uses the proposed metaheuristic algorithm in solving multilevel optimization problems.
PROMET - Traffic&Transportation, 2012
ABSTRACT
Many studies have been carried out using different metaheuristic algorithms on optimisation probl... more Many studies have been carried out using different metaheuristic algorithms on optimisation problems in various fields like engineering design, economics and routes planning. In the real world, resources and time are scarce. Thus the goals of optimisation algorithms are to optimise these available resources. Different metaheuristic algorithms are available. The firefly algorithm is one of the recent metaheuristic algorithms that is used in many applications; it is also modified and hybridised to improve its performance. In this paper, we compare the Standard Firefly Algorithm, the Elitist Firefly Algorithm, also called the Modified Firefly Algorithm with the Chaotic Firefly Algorithm, which embeds chaos maps in the Standard Firefly Algorithm. The Modified Firefly Algorithm differs from the Standard Firefly Algorithm in such a way that the global optimum solution at a particular iteration will not move randomly but in a direction that is chosen from randomly generated directions that...
Research Journal of Applied Sciences
Metaheuristic algorithms are useful in solving complex optimization problems. Genetic algorithm (... more Metaheuristic algorithms are useful in solving complex optimization problems. Genetic algorithm (GA) is one of the well known and oldest metaheuristic algorithms. It was introduced in 1975 and has been used in many applications varying from engineering to management and many other fields as well. However, Prey-Predator algorithm (PPA) is one of recently introduced algorithm, in 2012, inspired by the interaction between preys and their predator. The motivation and the search mechanism for these two algorithms are different. In this paper the comparison of these two algorithms both from theoretical aspects and using simulation on selected benchmark problems is presented. According to the results, PPA performs better than GA in the selected test problems. | Genetic algorithm (GA) | Prey-Predator algorithm (PPA) | Metaheuristic algorithms | Optimization |
Satisfiability of logic programming based on radial basis function neural networks
ABSTRACT In this paper, we propose a new technique to test the Satisfiability of propositional lo... more ABSTRACT In this paper, we propose a new technique to test the Satisfiability of propositional logic programming and quantified Boolean formula problem in radial basis function neural networks. For this purpose, we built radial basis function neural networks to represent the proportional logic which has exactly three variables in each clause. We used the Prey-predator algorithm to calculate the output weights of the neural networks, while the K-means clustering algorithm is used to determine the hidden parameters (the centers and the widths). Mean of the sum squared error function is used to measure the activity of the two algorithms. We applied the developed technique with the recurrent radial basis function neural networks to represent the quantified Boolean formulas. The new technique can be applied to solve many applications such as electronic circuits and NP-complete problems.
Nature-inspired optimization algorithms have become useful in solving di±cult optimization proble... more Nature-inspired optimization algorithms have become useful in solving di±cult optimization problems in di®erent disciplines. Since the introduction of evolutionary algorithms several studies have been conducted on the development of metaheuristic optimization algorithms. Most of these algorithms are inspired by biological phenomenon. In this paper, we introduce a new algorithm inspired by prey-predator interaction of animals. In the algorithm randomly generated solutions are assigned as a predator and preys depending on their performance on the objective function. Their performance can be expressed numerically and is called the survival value. A prey will run towards the pack of preys with better surviving values and away from the predator. The predator chases the prey with the smallest survival value. However, the best prey or the prey with the best survival value performs a local search. Hence the best prey focuses fully on exploitation while the other solution members focus on the exploration of the solution space. The algorithm is tested on selected well-known test problems and a comparison is also done between our algorithm, genetic algorithm and particle swarm optimization. From the simulation result, it is shown that on the selected test problems prey-predator algorithm performs better in achieving the optimal value.
Multilevel optimization problems deals with mathematical programming problems whose feasible set ... more Multilevel optimization problems deals with mathematical programming problems whose feasible set is implicitly determined by a sequence of nested optimization problems. These kind of problems are common in different applications where there is a hierarchy of decision makers exists. Solving such problems has been a challenge especially when they are non linear and non convex. In this paper we introduce a new algorithm, inspired by natural adaptation, using (1+1)-evolutionary strategy iteratively. Suppose there are k level optimization problem. First, the leader's level will be solved alone for all the variables under all the constraint set. Then that solution will adapt itself according to the objective function in each level going through all the levels down. When a particular level's optimization problem is solved the solution will be adapted the level's variable while the other variables remain being a fixed parameter. This updating process of the solution continues until a stopping criterion is met. Bilevel and trilevel optimization problems are used to show how the algorithm works. From the simulation result on the two problems, it is shown that it is promising to uses the proposed metaheuristic algorithm in solving multilevel optimization problems.
Solving vector optimisation entails the conflict among component objectives. The best solution de... more Solving vector optimisation entails the conflict among component objectives. The best solution depends on the preference of the decision-maker. Firefly algorithm is one of the recently proposed metaheuristic algorithms for optimisation problems. In this paper, the random movement of the brighter firefly is modified by using (1 + 1)-evolutionary strategy to identify the direction in which the brightness increases. We also show how to generate a dynamic weight for each component of the vector by using a fuzzy trade-off preference. This dynamic weight will be imbedded in computing the intensity of light of fireflies in the algorithm. From the simulation results, it is shown that using fuzzy preference is promising to obtain solutions according to the given fuzzy preference. Furthermore, simulation results show that the evolutionary strategy based firefly algorithm performs better than the ordinary firefly algorithm.
Firefly algorithm is one of the new metaheuristic algorithms for optimization problems. The algor... more Firefly algorithm is one of the new metaheuristic algorithms for optimization problems. The algorithm is inspired by the flashing behavior of fireflies. In the algorithm, randomly generated solutions will be considered as fireflies, and brightness is assigned depending on their performance on the objective function. One of the rules used to construct the algorithm is, a firefly will be attracted to a brighter firefly, and if there is no brighter firefly, it will move randomly. In this paper we modify this random movement of the brighter firefly by generating random directions in order to determine the best direction in which the brightness increases. If such a direction is not generated, it will remain in its current position. Furthermore the assignment of attractiveness is modified in such a way that the effect of the objective function is magnified. From the simulation result it is shown that the modified firefly algorithm performs better than the standard one in finding the best solution with smaller CPU time.
Transportation plays a vital role in the development of a country and the car is the most commonl... more Transportation plays a vital role in the development of a country and the car is the most commonly used means. However, in third world countries long waiting time for public buses is a common problem, especially when people need to switch buses. The problem becomes critical when one considers buses joining different villages and cities. Theoretically this problem can be solved by assigning more buses on the route, which is not possible due to economical problem. Another option is to schedule the buses so that customers who want to switch buses at junction cities need not have to wait long. This paper discusses how to model single frequency routes bus timetabling as a fuzzy multiobjective optimization problem and how to solve it using preference-based genetic algorithm by assigning appropriate fuzzy preference to the need of the customers. The idea will be elaborated with an example.
"Most real-life optimisation problems involve multiple objective functions.
Finding a solution ... more "Most real-life optimisation problems involve multiple objective functions.
Finding a solution that satisfies the decision-maker is very difficult owing to conflict
between the objectives. Furthermore, the solution depends on the decision-maker’s
preference. Metaheuristic solution methods have become common tools to solve these
problems. The task of obtaining solutions that take account of a decision-maker’s
preference is at the forefront of current research. It is also possible to have multiple
decision-makers with different preferences and with different decision-making powers. It
may not be easy to express a preference using crisp numbers. In this study, the preferences
of multiple decision-makers were simulated and a solution based on a genetic algorithm
was developed to solve multi-objective optimisation problems. The preferences were
collected as fuzzy conditional trade-offs and they were updated while running the
algorithm interactively with the decision-makers. The proposed method was tested using
well-known benchmark problems. The solutions were found to converge around the
Pareto front of the problems."
Nawaf Hamadneh, Saratha Sathasivam, Surafel Luleseged Tilahun and Ong Hong Choon
scialert.net
Abstract: Neural-symbolic systems are based on both logic programming and artificial neural netwo... more Abstract: Neural-symbolic systems are based on both logic programming and artificial neural network s. A neural network is a black box that clearly learns the internal relations of unknown systems. Radial Basis Function Neural Network (RBFNN) is a commonly-used ...
internationalconference.com.my
It is common to face multiple and usually conflicting objectives in a decision making process. Ho... more It is common to face multiple and usually conflicting objectives in a decision making process. However due to the conflict of these objectives a decision maker needs to give a preference among the objectives to aid the decision analyst in the process of optimizing and analyzing the decision process. The mainly used ways to express the preference in multiobjective optimization are by ordering the objectives according to their importance or giving weight for each objective and a binary comparison between the objectives. Most of the papers discuss these preferences using crisp numbers. However, it is logical and natural for the decision maker to express the preference fuzzily. This paper discusses the fuzzy way of representing the preference of the decision maker and its analysis and properties. Furthermore the idea will be extended for the case of multilevel optimization and for the case when there are multiple decision makers, where multilevel optimization is an optimization problem in which there are multiple decision makers hierarchically with their own objective function and controlling part of the variables.
Far East Journal of Mathematical Sciences ( …, Jan 1, 2011
Population based solution methods have become more common in solving multiobjective optimization ... more Population based solution methods have become more common in solving multiobjective optimization problems. These methods overcome the limitation of classical methods of finding a single solution with a single run of the algorithms. The studies of incorporating the decision maker's preference to genetic algorithm had done so far use binary importance comparison of objectives, which lacks uniformity. This study shows how to incorporate the fuzzy preference of a decision maker in evolutionary algorithm so that the solutions will zoom to the region in which the preference of the decision maker lies. Cumulative fuzzy tradeoffs are expressed using an appropriate probability distribution function which agrees with the fuzzy membership function. It will then be incorporated in the fitness evaluation stage of genetic algorithm. From the simulation result on selected test functions and comparing this with previous studies, it is shown that incorporating the decision maker's fuzzy preference will give solutions which satisfy the decision maker's preference better.
Fuzzy Preference Incorporated Evolutionary Algorithm for Multiobjective Optimization
Multiobjective evolutionary method is a way to overcome the limitation of the classical methods, ... more Multiobjective evolutionary method is a way to overcome the limitation of the classical methods, by finding multiple solutions within a single run of the solution procedure. The aim of having a solution method for multiobjective optimization problem is to help the decision maker in getting the best solution. Usually the decision maker is not interested in a diverse set of Pareto optimal points. So, it is necessary to incorporate the decision maker’s preference so that the algorithm gives out alternative solutions around the decision maker’s preference. The problem in incorporating the decision maker’s preference is that the decision maker may not have a solid guide line in comparing tradeoffs of objectives. However, it is easy for the decision maker to compare in a fuzzy way. This paper discusses on incorporating a fuzzy tradeoffs in the evolutionary algorithm to zoom out the region where the decision maker’s preference lies. By using test functions it has shown that it is possible to give points in the region on the Pareto front where the decision maker’s interest lies.
Uploads
Papers by Surafel L. Tilahun
Finding a solution that satisfies the decision-maker is very difficult owing to conflict
between the objectives. Furthermore, the solution depends on the decision-maker’s
preference. Metaheuristic solution methods have become common tools to solve these
problems. The task of obtaining solutions that take account of a decision-maker’s
preference is at the forefront of current research. It is also possible to have multiple
decision-makers with different preferences and with different decision-making powers. It
may not be easy to express a preference using crisp numbers. In this study, the preferences
of multiple decision-makers were simulated and a solution based on a genetic algorithm
was developed to solve multi-objective optimisation problems. The preferences were
collected as fuzzy conditional trade-offs and they were updated while running the
algorithm interactively with the decision-makers. The proposed method was tested using
well-known benchmark problems. The solutions were found to converge around the
Pareto front of the problems."