Association rules and deep learning for cryptographic algorithm in privacy preserving data mining

Abstract

Security in database is important as it contains sensitized information. Data mining provides mechanisms for intermediate representation of data. Privacy preservation is vital due to the security aspects. Privacy denotes mechanisms for allowing the data to be accessed in secured manner. The basic idea is to protect the data that are sensitive from data miners such that it is not possible to pull out the sensitive data from database. Association rules are crucial idea to change the dataset from the original data set. Association rules with cryptographic techniques has been used. They also demonstrate the applicability by applying this algorithm on real life data sets. This research work proposed a well-organized privacy preservation data-mining scheme with data-mining perturbation merged approach. It uses the association rules with cryptography techniques. The paper also demonstrates how neural networks is being applied for predicting the medical dataset. The paper also provides scope on how deep convolution neural network can be applied for medical analysis.