Sharif Digital Repository

The traders exchange currencies in an online foreign exchange market (Forex), so they need to know ways help them to predict trend of market. There are some methods to forecast exchange rate based on experiment. However they can submit good signals, they are too late for exchange. Two things are important, they should be correct and on time. In this research, we try to submit a prediction by Adaptive Methods in which provides both of them. In first approach, we test 468 models of Artificial Neural Network to achieve the best; and in second approach, Genetic Algorithm and Swarm Intelligence are applied to training Artificial Neural Network. Finally, in addition to forecasting exchange rate,... 

In this research a monopolist is considered which its object is maximization of profit during a predetermined period. Optimal control theory is used as a tool for determining of optimal trend for control variables. Price, warranty length and production rate are considered as control variables and cumulative demand and production are considered as state variables. Learning effect in production side and demand side are entered in the model too. In the demand side, learning effect shows its influences with movement of market from diffusion phase to saturation phase. In diffusion phase the demanded quantity of productions increases by increase of cumulative demand, on the contrary in the... 

In this research, a new metaheuristic algorithm is developed for solving the Discrete Time- Resource Trade off Problem in the field of project scheduling.In this problem ,a project contains activities interrelated by finish-start type precedence constraints and each has a specified work content and can be performed in different combinations of duration and resource requirement.Since the problem is NP-hard , the Particle Swarm Optimization is adopted due to minimization of the makespan subject to precedence relations and a single renewable resource. Basically PSO is used to solve continous problems and discrete problems have just begun to be solved by the discrete PSO.In proposed method,a... 

Asset and liability management is a method for optimizing cash flows of input and output of a bank . The model for optimizing the balance between liability and assets of a bank is presented in this research. Bank asset liability management and planning can be considered at the same time. In tjis model , all assets and liabilities in the bank balance sheet are discussed under different objectives and requirements , such as legal and administrate requirements, market condition, risk rate etc  

A hierarchical production system in make-to-order environment is presented in this thesis. This production system is consisted of three different decision making levels. Maximum acceptance for production within customer specified time interval was done in the first stage, which is called order arriving stage, based on different client criteria, type of demanded product, and production system constraints. A new hierarchical decision making structure for accepting or rejecting arriving orders is proposed in this level. The output of this acceptance level is orders having maximum production system defined criteria, and minimum production cost along with order delivery time with minimum delay is... 

In Computer-Aided Process Planning, there are two ways of manufacturing called Variant Manufacturing and Generative Manufacturing. Variant Manufacturing is based on a data base from a part family for a project. The part family is made in a series of root sheets regarding the final form of the intended part. The data base is updated after the manufacturing of each part. In order to make the system intelligent, one can use the OPITZ coding system that specifies the geometrical form, geometrical details and part manufacturing operations in a 13-digit number. The aim of this research is to choose the best machining parameters according to time and cost, that, based on the basic principles of... 

In this thesis, a multi-stage project portfolio selection and scheduling problem in an uncertain environment is modeled, analyzed and solved. Moreover, we investigate the situations in which a project manager is confronted to a large number of projects to be selected and scheduled while the required resources of the projects are limited and renewable. The aim is to maximize objective function which is the net present value of the profit of the projects. To solve the mentioned problem, we introduce a multi-stage stochastic programming. For the reason that, in real problems, optimization problems that included all restraints are large scale and have computational complexity, these problems are... 

In the field of supply chain, management constantly seeks to adopt strategies for realizing its targeted objectives. The supply chain, therefore, includes integrated production-distribution processes with all effective parts where proper management requires a comprehensive and dynamic approach. It is necessary to meet required items at required time. This becomes more important with the supply systems providing items. On the other hand, the difficulty of supply chain management, decisions, and policy-making can be increased by a wide range of conditions such as changes in demand. Thus, a comprehensive approach is needed to take the present dynamics into account. The current paper develops a... 

In recent years in assembly line balancing area many researchers have focused on identifying and modeling real world situations in assembly lines in order to obtain more suitable model. As a result many generalized models have been developed in assembly line balancing area. However, designing efficient algorithms to solve these generalized problems has just begun. In this study it is aimed to design efficient algorithms to solve one of these generalized problems. Specificcally two hybrid genetic algorithms (GA) are proposed to solve assembly line balancing problem with sequence dependent setup times. Scheduling of tasks in the stations has been less considered in the literature of assembly... 

Due date assignment for customer orders has been studied in various production environments. In this thesis a production system is considered in which the orders of several customers are produced in a single lot because of the economy of scale. If a lot is completed before receiving customer orders, inventory carrying cost is incurred. On shortages, a due date is quoted to customers and lead time quotation cost is incurred. The due date may not be met because of (lead-time) uncertainty in production system. Hence if the order is delivered after its due date, tardiness cost is charged. The literature reviews show that such a model is not considered by other researchers. Two different problems... 

Optimal investment strategies are often derived in continuous time models, but have to be implemented in discrete time. It has been shown that in models with stochastic volatility or jumps; this could lead to significant utility loss, for an investor who utilizes ‘Derivatives’ in his/her portfolio. In this study, we determine the optimal investment strategies with discrete trading explicitly taken into account, through ‘Stochastic Dynamic Programming’. These strategies are in the form of optimal factor exposures for portfolio. The investor, then, needs to use sufficient non-redundant Derivatives in addition to the ‘Stock’ to gain the desired exposures in each point of state space he meet.... 

In the proposed model, it is attempted to obtain the optimum cost of activities concerning their relevant risks. I n doing so, the concept of risk based project value (RPV) has been employed to measure the expected earn of each individual activity in the presence of risks. Generally, in the model, the risks are divided into critical and non-critical risks. Critical risks cause termination of activities and non-critical risks that are not as severe as critical risks and lead to monetary lost. In the next step, the risk response plans are utilized to reduce the possible adverse effects of risks. These plans, also known as action plans, are the preventive or adaptive actions that aim to reduce... 

In this thesis a two-tier supply chain containing an upstream supplier and a downstream producer have been modeled. This model presents a replicator interaction in supply chain. In this model supplier has to choose between two different production strategies which provide two types of output-quality. It is obvious the method with better output-quality has more production cost and the one with lower output-quality has lower production cost. The strategy of producer is to choose whether test the production before process it in production line or not. By the tool of evolutionary game theory trends of changing in players' strategies have been modeled and Nash equilibrium points and the types of... 

Knowledge markets where knowledge transactions take place are today see increasingly expansion by influences of the Internet and the importance of knowledge. A competitive space, thus, has been placed among different enterprises step in knowledge markets. With no exception to this general rule, one of these markets covers online question – answer knowledge market. Google Answers is a proposed website in this area which, although being of interest in recent decades, its activities ended in 2006 for several reasons. Significantly, these markets consist of many factors with interrelated effects on each other and this leads to complexity of systems. Therefore, in order to study these markets, it... 

In this thesis, project portfolio selection under uncertainty conditions is considered. Firstly, the problem is modeled with mathematical programming and also this problem has two objective functions that contain maximizing profitability and minimizing surplus resources. Each project has resource costs and contract prices. In addition, uncertainties for these parameters should be considered. Secondly, another innovation in this research is considering Cost/Time trade-off for duration and cost. To satisfy this subject, each project has different scenarios with various duration and usage of resources. Since the project has two objective functions, ε-constraint method is applied to optimizing... 

Supply chain problems consist of a large number of issues. One of the key problem in this area is supplier selection in supply chain. Several criteria must be considered in the supplier selection. So, multi criteria decision making approach, neural network, mathematical programming and etc. are used for solving this problems. In most of these problems, supplier selection is done according to the buyer’s objectives.
So far, several methods are provided for supplier selection problem. In majority of them, supplier selection problem is surveyed from buyer’s viewpoint and in those have only been considered the buyer’s benefit maximization that will not necessarily lead to optimal condition... 

Blood is a perishable product and can only be stored for a limited time. Reducing the level of wastage and shortage in supply chain of blood products is the main challenge of supply chain management. Wastage occurs whenever hospitals and blood banks order excessive blood while ordering insufficient amount results in shortage. Wastage can increase the costs in supply chain while shortage can cause important operation to be cancelled.This thesis presents an integer programming model for improving blood related operations in hospitals and blood banks. The presented model determines optimal order levels for each period which minimize total cost, shortage and wastage levels. The model is... 

Crude oil is a key element in world economy and the most widely traded form of energy. Therefore, a clear and effective understanding of crude oil price behavior is of great importance for businesses, governments and policy makers. Taking into account the exhaustible nature of crude oil and impact of monetary policy along with other major drivers of crude oil prices, this paper investigates predictability of oil prices using artificial neural networks. A Multilayer Perceptron (MLP) neural network is developed and trained with historical data from 1980 to 2014 and using mean square error (MSE) for testing data, optimal number of hidden layer neurons is determined. Meanwhile, an economic model... 

Decision Tree is used to analyze situations with sequential decisions. Applying Decision tree method in such situations leads to identifying the best sequence of decisions. On the other hand, System Dynamics is a powerful method to gain useful insight into situations of complexity, interconnections, feedback loops nonlinearities and time delays. System Dynamics is able to model complexities and interconnections of situations and provide precise understanding about it. Therefore, combination of these two methods could be very effective in analysis of problems that firstly, comprise of sequential decisions with multiple uncertainty and secondly, illustrate complex behavior. Supply chain is one... 

Scheduling has been an active area of research for decades. A substantial amount of work has been done to define, classify, and solve scheduling problems. Most of these problems are computationally difficult to solve, and incorporating other decisions into them increases the complexity of these problems. This thesis focuses on order acceptance and scheduling (OAS) problems, and proposes time-indexed mixed integer and linear programming (MILP) models for the first time. An OAS problem can be defined with parameters of processing time, release date, due date, and deadline for each order. There are also a maximum revenue that will be brought by each order and an importance weight of each order....