Root Causes Prediction in Data Center Using Convolutional Dense Neural Network
Jeremy Filbert Baskoro, Fidelio Soares de Carvalho, Catur Apriono · 2024
Rapid advances in telecommunications, media, information technology and the widespread development of global information infrastructure have changed the patterns and ways of doing business in industry, commerce, and government. A reliable data center must support these industries. However, a data center with a complex network architecture can cause problems or system failures, such as database issues, memory, and network response. This research proposed a root cause prediction in a data center using a combination of CNN and DNN, considering the dataset provided by an open-source dataset. This research also considers zero padding and dropouts to avoid overfitting and adds convolutional layers to improve accuracy. The matrix confusion and the training accuracy are above 90% and up to 82 %, respectively. These results indicate that the proposed model can provide a reliable root cause prediction in a data center.