An Implementation of Anomaly Detection in IoT Medical Data using Deep anomaly Detection Models
K.V.Daya Sagar1, DBK Kamesh2

1K.V.Daya Sagar, Research Scholar, Shri Venkateshwara University, Gajraula, ( U.P), India.
2DBK Kamesh, Professor, Mall Reddy Engineering College for Women, Maisammaguda, Secunderabad, (U.P.), India.
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 578-582 | Volume-8 Issue-8, June 2019 | Retrieval Number: H6720068819/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Anomaly detection is a challenging task for IoT data, which aims to identify observations that deviate from a nominal sample. Traditional distance-based, mostly anomaly-based methods of detection estimate the neighborhood distance between each view and suffer in high space from the curse of dimensionality. In this paper, we propose a hybrid semi-supervised model of anomaly detection for high-dimensional data consisting of two parts: a deep autoencoder (DAE) and an anomaly detector based on the nearest neighbor graphs (NNG). Benefiting from nonlinear mapping ability, the DAE is first trained to learn the essential features of a high-dimensional dataset to represent high-dimensional data in a more compact subspace. Several nonparametric anomaly detectors based on KNN are building those randomly sampled from the entire dataset from different subsets. All detectors of anomaly make the final prediction. The proposed method is evaluated on the ECG data set generated by the IoT Sensing element node and the results also ensure that the proposed hybrid model improves the accuracy of detection and reduces the complexity of the computation.
Keyword: ECG, Wearable sensors, Anomaly, DAE, Deep autoencoder, NNG, KNN.
Scope of the Article: Distributed Mobile Applications Utilizing IoT.