Based on the findings of Massachusetts Institute of Technology, organizations’ data double every five years. However, the rate of using data is 0.3. Nowadays, data mining tools have greatly facilitated the process of knowledge extraction from a welter of data. Thi
More
Based on the findings of Massachusetts Institute of Technology, organizations’ data double every five years. However, the rate of using data is 0.3. Nowadays, data mining tools have greatly facilitated the process of knowledge extraction from a welter of data. This paper presents a hybrid model using data gathered from an ATM manufacturing company. The steps of the research are based on CRISP-DM. Therefore, based on the first step, business understanding, the company and its different units were studied. After business understanding, the data collected from sale's unit were prepared for preprocess. While preprocessing, data from some columns of dataset, based on their types and purpose of the research, were either categorized or coded. Then, the data have been inserted into Clementine software, which resulted in modeling and pattern discovery. The results clearly state that, the same Machines’ Code and the same customers in different provinces are struggling with significantly different Problems’ Code, that could be due to weather condition, culture of using ATMs, and likewise. Moreover, the same Machines’ Code and the same Problems’ Code, as well as differences in Technicians' expertise, seems to be some causes to significantly different Repair Time. This could be due to Technicians' training background level of their expertise and such. At last, the company can benefit from the outputs of this model in terms of its strategic decision-making.
Manuscript profile
