Anfis Modeling To Forecast Maintenance Cost of Associative Information Technology Services

The model in this study was created to apply a new and functional model for prediction of it maintenance cost of system downtime by using artificial intelligence method in contrast of other research techniques in it cost measuring. the estimation is based on the measuring of it services availability. in this model, anfis was developed for quantifying information technology generated services (itgs) perceptible by business users. in addition, it was used to forecasting it cost related to system maintenance that it can help managers for future and constructive decision. this model has been applied and tested prediction technique by anfis in matalb fuzzy toolbox by previous large volume of data gathering from it cost factors, itgs, and associative cost in order to building pattern, tuning and training this model well. first of all, the model was fully developed, stabilized, and passed through intensive training with large volume of data collected in one of the organization in iran. in first phase, it can be possible to feed a specific period of data into the model to determine the quantity of services (itgs), and in second phase, their related maintenance cost can be predicted. anfis learning technique predicted maintenance cost of measured services availability which it was totally provided with first quantifying services in a specific time period. having an operational mechanism for measuring and quantifying information technology services tangible by users for estimating their costs is contributed to practical accurate investment. some components have been considered and measured in the field of system maintenance. the main objective of this study is prediction the amount of investment for maintenance of entire itgs by extraction and considering the factors affecting the software maintenance cost help to estimate the cost and reduce it by controlling the factors.