during the process lead to significant losses to the plant. In this paper, in order to detect
defects and fault patterns, Support Vector Machine (SVM) which is one of the machine
learning algorithms, classifies the data received from the equipment as normal or abnormal.
After learning only normal data by using Restricted Boltzmann Machine (RBM). We propose
a model to identify the data, and then we analyze the faults of facilities in real-time.