Intelligent data deduplication with Deep Transfer Learning Enabled Classification Model for Cloud-based Healthcare System

Abstract

The extensive adoption of cloud computing (CC) enabled healthcare systems in attaining medical data from different data sources sustained by heterogeneous cloud providers. Data deduplication (DD) is a proficient method to effectually share and store data in the cloud server. At the same time, deep learning (DL) models can be applied for effective decision-making in the cloud-based healthcare system. With this motivation, this study develops an Intelligent DD with Deep Transfer Learning Enabled Classification Model for Cloud-based Healthcare System (IDDTLC-CHS) model. The presented IDDTLC-CHS model aims to accomplish effective DD and classification in the cloud-enabled healthcare environment. For DD, the neighbourhood correlation sequence (NCS) algorithm is employed which generates optimum code words and is then compressed by Deflate model. Besides, the data classification module involves a series of processes namely fuzzy c-means (FCM) segmentation, Xception feature extraction; bidirectional gated recurrent unit (BiGRU), and seagull optimization algorithm (SGO). The SGO algorithm is applied for optimal adjustment of the parameters involved in the BiGRU model. To assess enhanced outcomes of the IDDTLC-CHS model, a wide-ranging simulation analysis is carried out using the benchmark dataset. The comparative analysis reported the betterment of the IDDTLC-CHS model compared to other recent approaches.