دانلود مقاله ISI انگلیسی شماره 28504
عنوان فارسی مقاله

شرح مختصر ژنراتور هوشمند ترکیبی برای برنامه ریزی استراتژیک کسب و کار با استفاده از ANFIS(سیستم استنتاج عصبی-فازی تطبیقی )

کد مقاله سال انتشار مقاله انگلیسی ترجمه فارسی تعداد کلمات
28504 2009 9 صفحه PDF سفارش دهید 5550 کلمه
خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.
عنوان انگلیسی
Hybrid intelligent scenario generator for business strategic planning by using ANFIS
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Expert Systems with Applications, Volume 36, Issue 4, May 2009, Pages 7729–7737

کلمات کلیدی
شرح مختصر برنامه ریزی - سیستم عصبی فازی - سیستم استنتاج عصبی - فازی تطبیقی
پیش نمایش مقاله
پیش نمایش مقاله شرح مختصر ژنراتور هوشمند ترکیبی برای برنامه ریزی استراتژیک کسب و کار با استفاده از ANFIS(سیستم استنتاج عصبی-فازی تطبیقی )

چکیده انگلیسی

The aim of this study is to investigate a new method for generating scenarios in order to cope with the data shortage and linguistic expression of experts in scenario planning. The proposed hybrid intelligent scenario generator uses an Adaptive Neuro-Fuzzy Inference System (ANFIS) to deal with uncertain inputs. In this methodology, the strengths of expert systems, fuzzy logic and Artificial Neural Networks (ANNs) are joined to generate possible future scenarios. The proposed methodology includes four steps: step 1 defines the scope and internal and external variables and step 2 determines rules from experts. Then, step 3 prepares ANFIS system which is conducted by computer programming in Matlab environment. The Last step is sensitivity analysis to study the effects of variation of inputs on outputs. The applicability of the proposed method has been tested against two different case studies.

مقدمه انگلیسی

The purpose of strategic planning is to guide an organisation to achieve its desired goals of the long-term development under the variation of environment (Wang, 1999). Therefore, the future events play a key role in business strategic planning and managers need a mental model of the future to make better decisions. There are some differences among uncertainties pertaining to future occurrence probability. When there is the low level of uncertainties in environment, quantitative approaches such as probability distribution and forecasting techniques are very useful for managing the existing risk and uncertainty. In the high level of uncertainty, qualitative approaches such as scenario planning may be useful to employ (Alessandri, Ford, Lander, Leggio, & Taylor, 2004). Scenario planning is not aimed at obtaining a forecast but instead produces alternative images of the future which can avoid the pitfalls of more traditional methods (Goodwin and Wright, 2001 and Postma and Lieb, 2005). Managers are able to have much better positioning with regard to unexpected events by using scenario planning methods. Scenario planning attempts to capture the richness and range of possibilities, and considers changes that decision makers would otherwise ignore (Schoemaker, 1995). Scenario planning has been defined as “a process of positioning several informed, plausible and imaginative alternative future environment in which decisions about the future may be played out for the purpose of changing current thinking, improving decision making, enhancing human and organizational learning and improving performance” (Chermack & Lynham, 2002). Various scenario planning approaches from literature are classified into two major categories: qualitative and quantitative. SRI (Ringland, 1998), Future Group (Chermack, Lynham, & Ruona, 2001), Global Business Network (Chermack et al., 2001), Schoemaker (Schoemaker, 1995), and DSLP (Royes & Royes, 2004) methodologies are all subjective, qualitative in nature and firmly process-oriented. This means that organisational learning process in these approaches is more important than the reliability of the content of the end product, which is the scenarios (Bradfielda, Wrightb, Burta, Cairnsb, & Van Der Heijdena, 2005). These approaches are not based on past data but consider qualitative and subjective information of experts to construct scenarios. On the other hand, Godet’s methodology (Godet, 2001, Godet, 2006 and Godet and Roubelat, 1996) which has been known as a quantitative method is essentially outcome-oriented. A quantitative methodology develops scenarios for particular phenomenon and sets key variables for a specific subject. The experts’ rules in quantitative methodologies are dominant and they judge about the occurrence probability of each scenario. Quantitative methodologies such as Godet’s framework consider the conditional probability of each occurrence which is assumed for different sets of environmental and organisational variables. In all scenario planning methodologies, experts’ role is critical for decision making, and uncertain data always are the basis for developing future scenarios. Scenario planning deals with uncertain and ambiguous data and therefore, some researchers applied fuzzy logic and Artificial Neural Networks (ANNs) for better handling of the data shortage and also experts’ linguistic expression. Khoo, Ho, and Choa (1994) developed a fuzzy management decision support system for scenario analysis using a hybrid technique: a combination of the fuzzy Delphi analysis and fuzzy reasoning technique. Wang (1999) proposed a method of fuzzy scenario analysis to forecast the possible development in a strategic planning. This method considered the uncertainties involved in strategic planning to determine the compatible and possible scenarios. Li, Ang, and Gay (1997) developed a scenario generation tool by using the theory of ANNs and truth value flow inference. ANNs were designed to forecast market share and market growth, and a fuzzy expert system model was developed to build a knowledge-base for defining a suitable marketing strategy. Royes and Royes (2004) developed a framework to indicate how the fuzzy set approach may contribute to the evaluation and exploration of scenarios for strategic planning. A hybrid methodology was developed which used three main modules: fuzzy sets, multicriteria analysis and case-based reasoning. This paper presents a new hybrid methodology to combine the advantages of fuzzy logic and ANNs. Other methodologies use only one method, fuzzy logic or ANNs, for dealing with the data shortage and experts’ linguistic expression. Li’s method is the only methodology developing a hybrid intelligent system based on fuzzy logic and ANNs but the major difference of this method with the proposed methodology is related to the proposal using ANNs. In Li’s framework, ANNs are utilized for forecasting market share and growth, while the suggested methodology applies ANNs as a tool to learn from experts and make decisions. The main goals of this new hybrid intelligent architecture will be: – To improve the ability of managers to deal with uncertainty. – To present intelligent advice on business strategic planning. – To keep and use the experts’ knowledge. To attain these objectives, the proposed framework applies an Adaptive Neuro-Fuzzy Inference System (ANFIS) which is suggested by Jang (1993) to better deal with an ill-defined and uncertain system. It can serve as a basis for constructing a set of fuzzy if–then rules with appropriate membership functions to generate the stipulated input–output pairs (Jang, Sun, & Mizutani, 1997). ANFIS architecture is designed to tune fuzzy system parameters based on input/output pairs of data. The fuzzy inference process is implemented as a generalised ANN, which is then tuned by gradient descent techniques (Fuller, 2000). Antecedent parameters of fuzzy rules are also tuned as well as consequent parameters. ANFIS provides a method for the fuzzy modelling procedure to learn information about a dataset, in order to compute the membership function parameters allowing the associated fuzzy inference system to track the given input/output data (Huang, Chen, & Huang, 2007). ANFIS is used in many areas such as forecasting (Aznarte et al., 2007), classifying (Ozturk et al., 2008 and Sengur, 2008), controlling (Elmas, Ustun, & Sayan, 2008), recognition (Avci and Avci, 2007 and Avci et al., 2007) and diagnosing (Güler and Übeyli, 2004, Polat and Gunes, 2007 and Übeyli, 2008). The goal of this research is to develop an intelligent scenario generator based on ANFIS to eliminate the weaknesses of previous methodologies. The theory of ANNs will be used to enable learning and correcting from experts. Furthermore, fuzzy logic theory will be applied to deal with reasoning and using linguistic information acquired from experts. In Section 2, the details of ANFIS methodology for generating scenarios have been described. Two case studies are introduced in Section 3. The findings of case studies will be discussed in Section 4. In final section, the result of this research will be explained.

نتیجه گیری انگلیسی

The development of a hybrid intelligent scenario generator has been described in this paper. The hybrid architecture is the central point of the proposed methodology which allows having fuzzy rules and a learning algorithm in scenario generation. ANNs bring out the ability to learn from experts and fuzzy logic gives a consensus to express the ambiguity in human thinking and is able to mimic the human reasoning process. Therefore the developed methodology has the ability to learn and correct experts and also translate the linguistic experts’ rules. Moreover, ANFIS combines the classical back propagation methods with the minimum least square methods to modify the fuzzy numbers for uncertain variables. In this paper, two case studies are also presented to demonstrate the applicability of the hybrid intelligence scenario generator. The results of case studies show that the inputs greatly affect the selected strategic option and highlight the need for an optimisation strategy for finding the best strategic option.

خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.