hamed samarghandi

An accounting information system is a platform in which the basis of accounting is formed and managers and all users make decisions based on its product. Playing such a role necessitates the accurate identification of the actors playing a role in this system in order to plan in a principled and principled way in the development and productivity of the country's financial and economic system. Therefore, if it is possible to accurately predict and identify the actors affecting the accounting information system before the start of activities, a fundamental change in the field of management of the accounting information system can be created. The purpose of this study was to predict the prioritization of non-human actors affecting the accounting information system in Iran from the perspective of the network of actors using four algorithms: support vector machine (SVC), random forest (RF) ٬ Boosting and K-nearest neighbor (k- NN) is. Findings show that the most important actors in the accounting information system are organizational actors, followed by technical actors, economic actors, technology actors, and socio-political actors who have the most predictive power in the accounting information system. The RF model is also known as the best model for predicting the most effective characteristics of non-human actors in the accounting information system.

The accounting software is considered to be of the most critical components of accounting information system, with particular significance as of accounting and financial systems. the most important problems with accounting education systems is that students do not adequately learn the financial software required by the accounting profession, which, in turn, reduces the credibility and position of the accounting profession. That the main objective of accounting software education is to educate skilled and expert accountants to enter the accounting profession, which is considered as of the success factors of country’s economy. In this study, employ data mining techniques to investigate the accuracy, precision, and recall performance measures and to predict the rate of financial software learning based on accounting students’ emotional intelligence (EI), gender and education level. Accordingly, a machine-learning-based multivariate statistical analysis is performed on 100 Iranian accounting students. The results show that emotional intelligence has the most impact on the rate of financial software learning among the variables. Gender and education level were influential. Also, among the five algorithms, the highest precision and recall are achieved by both Decision Tree and XGBoost and are presented as the most appropriate models for the prediction rate of financial software learning.

The impetus for this research was examining a flow shop problem in which tasks were expected to be successively carried out with no time interval (i.e., no wait time) between them. For this reason, they should be completed by specific dates or deadlines. In this regard, the efficiency of the models was evaluated based on makespan. To solve the NP-Hard problem, we developed two mathematical models. Once we solved our problem using Mixed-Integer Programming Model (henceforth MIPM) and then, we applied a Constraint Programming Model (CPM); finally, we compared the optimality of the presented results.