سلیمان منصوری محمدآبادی

Performance evaluation is one of the most important measures required to improve and promote productivity. In this regard, determining an efficient method for evaluation appropriate to the conditions is essential. Data envelopment analysis is a non-parametric mathematical method for evaluating decision-making units. Several models have been proposed in order to evaluate performance with the data envelopment analysis method, such as the basic CCR, BCC, and Anderson-Peterson models. After the basic models, new models such as network data envelopment analysis models and ... were proposed due to the shortcomings of the basic models. The existence of shortcomings such as the multiplicity of zero weights, lack of model integration, lack of separability, uncertainty in the data, etc. led us to develop the basic CCR model in this research. The main basis of the new models presented are issues such as model integration with the aim of the greatest distance from non-ideal inputs and the least distance from ideal outputs, as well as considering constraints such as the smaller efficiency constraint equal to one, the equality constraint by considering an ideal unit with the greatest output and the least input, in some models, normalizing the values ??of inputs and outputs, considering the upper and lower limits for the weights of inputs and outputs. In this regard, pharmaceutical companies on the Tehran Stock Exchange were selected as a case study in order to implement the designed models. In this research, three methods were proposed and implemented according to the type of data. For deterministic data, the deterministic model was used, for uncertain data, the robust model of Bertsimas and Sim (۲۰۰۴) and the envelopment analysis model were combined, and for uncertain data where the dispersion distribution is uncertain, we used the combination of data envelopment analysis with FCM-KDE. In the present study, based on previous research and expert opinions, we have used p/e, p/b, asset turnover ratio, current ratio, return on equity, return on assets, and ROA as appropriate indicators to evaluate the performance of pharmaceutical companies in the stock market industry. According to the prioritization, Depars, Daravar, Decapsul, Deshimi, and Defara companies have been assigned the first to fifth priorities, respectively. Also, the decision-making units of Dezahrawi, Druz, Desasho, Diran, and Deabid have been assigned the ۵th priority from the last. Other decision-making units have been assigned the sixth to twenty-third priorities. The correlation coefficient of the indicators indicates acceptable correlation of the models in all cases. Of course, in the multi-faceted, deterministic and stable state, the data are almost similar and close to each other, but in the data-driven state, the data are not necessarily similar and close to the deterministic data; In this way, the data consists of ۳۰ clusters, each of which has a different chance of occurrence. Given that in the data-driven mode, the determined clusters express different states of the probability of occurrence of the parameter and the data is closer to the real world; this method is recommended in calculating the performance of companies.