Gulgun Kayakutlu

Using artificial neural network models in stock market index prediction

Forecasting stock exchange rates is an important financial problem that is receiving increasing attention. During the last few years, a number of neural network models and hybrid models have been proposed for obtaining accurate prediction results, in an attempt to outperform the traditional linear and nonlinear approaches. This paper evaluates the effectiveness of neural network models which are known to be dynamic and effective in stock-market predictions. The models analysed are multi-layer perceptron (MLP), dynamic artificial neural network (DAN2) and the hybrid neural networks which use generalized autoregressive conditional heteroscedasticity (GARCH) to extract new input variables. The comparison for each model is done in two view points: Mean Square Error (MSE) and Mean Absolute Deviate (MAD) using real exchange daily rate values of NASDAQ Stock Exchange index.

Developing Oregon's renewable energy portfolio using fuzzy goal programming model

Renewable energy continues to be a hot topic in the United States affecting security and sustainability. A model to create renewable energy portfolio is established using guidelines drawn by Oregons Renewable Portfolio Standard (RPS) legislation with the objective of responding to a 25% of the state electricity demand by renewable resources in 2025. The fuzzy goal programming model is adaptable to accommodate changes in energy costs and future advances in technology maturity. It can also take into consideration the preferences of policy-makers and stakeholders. This model can help to reveal the costs and benefits of complex decisions regarding renewable energy.

Scenario analysis using Bayesian networks: A case study in energy sector

This paper provides a general overview of creating scenarios for energy policies using Bayesian Network (BN) models. BN is a useful tool to analyze the complex structures, which allows observation of the current structure and basic consequences of any strategic change. This research will propose a decision model that will support the researchers in forecasting and scenario analysis fields. The proposed model will be implemented in a case study for Turkey. The choice of the case is based on complexities of a renewable energy resource rich country. Turkey is a heavy energy importer discussing new investments. Domestic resources could be evaluated under different scenarios aiming the sustainability. Achievements of this study will open a new vision for the decision makers in energy sector.

Effective supply value chain based on competence success

Purpose – This paper seeks to propose a managerial decision framework for different levels of supply chain, by addressing the strategic importance of competence values in supply chain effectiveness.Design/methodology/approach – A conceptual framework for supply chain effectiveness is defined in levels of supply chain targets, knowledge management dynamics, competence levels and competence success attributes. Analysis of literature in the areas of competence management, knowledge management, supply chain and value chain management resulted in defining the factors of the model. Surveys of industrial practices were used to validate the choice of factors. The analytical network process (ANP) is used to determine the most beneficial competence success attributes in a case study performed for three companies that participate in different stages of the textile supply chain.Findings – Individual competence in continuous learning and networking, as well as innovativeness of the team are found to be the three most ...

Assessment of lean manufacturing effect on business performance using Bayesian Belief Networks

This study aims to analyze lean manufacturing effect on financial, non-financial and sustainability performances.Causal maps are used for defining the inter-relation and intra-relation of lean factors.Bayesian Belief Networks are used for linking the lean factors.Effects of various combinations of lean factors on business performance are analyzed.Developed scenarios can be a lead for defining the business strategies based on lean thinking in any industry. The challenge of agility for adopting new business norms creates the need for measuring performance under changing conditions. This study aims to demonstrate the financial and non-financial consequences of implementing different combinations of lean techniques on the business performance. Bayesian Belief Network is used in studying the effects of factors under changing conditions. There are seven lean factors and four achievements studied to analyze the impact on three performance indicators. Bayesian Belief Network is constructed on the lean aspects that stimuli flexibility, reliability, quality and time of operations, which will have positive impacts on the financial, non-financial and sustainability performances of suppliers. A case study is carried out for suppliers in the automotive industry and scenarios with different combinations of lean factors are studied. This study gives a new vision in applying Bayesian network for business performance measures considering both the tangible and intangible results under changing business conditions.

Identification of energy policy priorities from existing energy portfolios using hierarchical decision model and goal programming : case of Germany and France

Purpose – There has not been a technology assessment study conducted that can be found publicly that created the energy policies of France and Germany – which are two nations that represent the vast extremes of national energy mix in Europe. This paper aims to use objective data to reverse generate the energy policy priorities that would yield the current portfolios.Design/methodology/approach – A hierarchical decision model was developed and analyzed backward using the goal programming to identify the priorities of criteria in a variety of technology alternatives. The three levels of goals are combined in one discrepancy minimization. Goals are defined in a non‐linear way to determine the values in different levels. Although the goals are given the same priority, the proposed model is used to determine the weight of each criterion and then finds the weight of technical, political/social and economic influences. Discrepancies achieved after having run the goal‐programming model determines the unbalanced p...

Value stream maps for industrial energy efficiency

Lean thinking is an engineering approach to avoid non-value adding tasks or processes in manufacturing. Energy is used by manufacturing companies for direct production processes, space conditioning and facility support. Most of the lean analysis of energy studies are multivariable statistics application focused on energy use in processes. This paper aims to reduce energy utilization by the small and medium manufacturing companies by implementing energy based Value Stream Mapping. Value Stream Mapping is a graphical technique allows to detect the level of value added by a process. This study adopts Value Stream Mapping technique to detect level of energy use at each step of different processes either in production or in facility support. The determined energy utilization level is validated in accordance with the outside temperature, production volume and the amount of solar effect on the facility. This will alow the discovery of energy saving opportunities. The suggested model can be used not only for diagnostic purposes but also for energy budgeting and saving measures. A case study application is given to demonstrate the Energy Value Stream Maps (E-VSM).

Forecasting Stock Exchange Movements Using Artificial Neural Network Models and Hybrid Models

Forecasting stock exchange rates is an important financial problem that is receiving increasing attention. During the last few years, a number of neural network models and hybrid models have been proposed for obtaining accurate prediction results, in an attempt to outperform the traditional linear and nonlinear approaches. This paper evaluates the effectiveness of neural network models; recurrent neural network (RNN), dynamic artificial neural network (DAN2) and the hybrid neural networks which use generalized autoregressive conditional heteroscedasticity (GARCH) and exponential generalized autoregressive conditional heteroscedasticity (EGARCH) to extract new input variables. The comparison for each model is done in two view points: MSE and MAD using real exchange daily rate values of Istanbul Stock Exchange (ISE) index XU10).

Patent value analysis using support vector machines

Receiving patents or licenses is an inevitable act of research in order to protect new ideas leading innovation. Request for patents has increased exponentially in order to legalize the intellectual property. Measuring economical value of each patent has been widely studied in the literature. Majority of the research in this field is focused on the patent driver prospect handled for the patent offices. There are a variety of criteria affecting decisions on each patent right; and predicting the possibility of grant may help the researchers to take some precautions. Objective of this study is to propose a robust model to determine if the appeal has a chance of approval. A case study is run on the patents that are accepted and rejected in home appliance industry to construct an intelligent classification model. The support vector machine, Back-Propagation Network and Bayes classification methods are compared on the proposed model. The proposed model in this study will help the decision makers to predict whether the patent appeal will be accepted. The study is unique with the approach that helps the candidate patent owners.

Clean energy investment scenarios using the Bayesian network

Clean energy investment decisions are getting more difficult to make due to public reactions. In order to support the policies in the field, analysis of the positive conditions is needed. This research aims to construct the positive scenarios for nuclear energy and renewable energy investments in the state of Oregon, USA. The Bayesian network technique will be used to create the scenarios. Oregon has a wide range of renewable energies; hence, investment is becoming more complex. Criteria affecting the decisions are taken from the literature, but were reviewed with energy authorities in Oregon in order to define the interactions.