An Approach to Identify the Missing Patent in Healthcare using the VAE Deep Learning Model- An Initial Move in Big Data Analytics.

Abstract

After having established a thorough comprehension of data, we also focus on the modeling that incorporates some present-day techniques such as that of variational autoencoders (VAE), which is one of the generative models that has been proved to work efficiently on most complex distributions by deep learning and variational inference. Additionally, the authors present a novel statistical learning approach that fuses several latent variable models and seeks to solve the major bottleneck with the current models – the reconstruction error, which is essential for the bridging the model. Such an approach does work to manage both 6.8% and 18% of youths but however portrays youths making use of patent fillip only have biometric modalities to do with the person to be patent filled. The effectiveness of the VAE method as proposed was tested against eleven other methods so as to ensure that standards of the technique were met by mean square error. For that specific purpose, the filled in data sources are uploaded to HDFS in order to provide analytics over the filled-in data. The ‘Heart Failure Clinical Records Dataset’ (HFCRD) from UCI repository contains both, bivariate and univariate data types. The data sub-set comes with differential percentile distribution on the degree of missingness.

The accuracy of understanding the imputation method used in the proposed work is about 87% which is the most in the imputation approach therefore the operable value created is other.