Medical steel fault prediction using deep learning techniques

Fault detection and analysis is considered to be an important factor in industrial production for medical applications. As industrial era has evolved tons, new fault identification ways are required to differentiate faults with totally good distinctions. The superior best a production is wanted to possess, the higher fault identification methodology the factories should apply. This research work focus on the assessment and evaluation of fault detection using deep learning techniques. The evaluation is made accordingly using Deep CNN with the variants corresponding to simple CNN, Resnet, Alexnet and Vgg_16. Besides, classification accuracy is improved by parameter optimizing and sample size equalization strategy. Experimental results shows that evaluation using the proposed methods with Vgg_16 gives an improved training accuracy of about 90% and validation accuracy of about 87%. This proves that fault detection and analysis in medical equipments and transplanting devices can be efficiently identified for better treatment and device management.

---

1. Track 3: Advances in Information Systems and Technology

---

2. Communication papers