A Novel Approach of Sensitive Data Classification using Convolution Neural Network and Logistic Regression
Gitanjali1, Kamlesh Lakhwani2
1Gitanjali , Professional University Phagwara Punjab India in department of Computer Engineering.
2Kamlesh Lakhwani, Associate Professor, in Lovely Professional University Phagwara Punjab India in department of Computer Engineering.
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 2883-2886 | Volume-8 Issue-8, June 2019 | Retrieval Number: H6862068819/19©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 (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Text Classification is a basic approach of text mining and natural language processing. In previous use, classifiers use human interface features like frequency base and n-gram features which are not able to find non-linearity in features and increase overlapping in features which directly impacts the performance of classifiers. In this paper, proposed convolution-based approach refines the traditional features in layered approach by activation function. This process increases the effective pattern for learning which is learn by Logistic regression and optimized by boosting approach. In experiment, there is comparison of machine learning approach which uses traditional features and deep learning approach which refine the traditional approach for increasing non-linearity pattern. The results showed that proposed approach CNN-Logistic regression improves the accuracy significantly because of the improving pattern of features.
Keyword: Text, Classification.
Scope of the Article: Classification.