Journal Title: International Journal of Science Engineering and Advance Technology - Year 2017, Vol 5, Issue 8
Abstract
We examine one profound learning technique named stacked denoising autoencoder (SDA). SDA stacks a few denoising autoencoders and connects the yield of each layer as the learned portrayal. Each denoising autoencoder in SDA is prepared to recoup the information from a ruined form of it. We build up another content portrayal display in view of a variation of SDA: marginalized stacked denoising autoencoders (mSDA), which receives straight rather than nonlinear projection to quicken preparing and minimizes limitless commotion dissemination keeping in mind the end goal to take in more vigorous portrayals. We use semantic data to grow mSDA and create Semantic-upgraded Marginalized Stacked Denoising Autoencoders (smSDA). The semantic data comprises of bullying words.
Authors and Affiliations
G. V. Chakradhar| Final M.Tech Student, Dept of Computer Science and Engineering, Prasiddha College of Engineering and Technology, Anathavaram-Amalapuram-533222, E.g.dt, A.P., V Swamy Naidu| Asst.Professor, Dept of Computer Science and Engineering, Prasiddha College of Engineering and Technology, Anathavaram-Amalapuram-533222, E.g.dt, A.P., M. Veerabhadra Rao| Head of the Department, Dept of Computer Science and Engineering, Prasiddha College of Engineering and Technology, Anathavaram-Amalapuram-533222, E.g.dt, A.P.
Keywords
Related Articles
- EP ID EP17006
- DOI -
- Views 227
- Downloads 8
Spontaneous Assessment Packet Origination
Nowadays systems aren't getting any slighter, they are expanding in size and it's transforming into a monotonous employment for system chairmen to adjust the system since they put trust in conventional instruments li...
Distributed Manipulation Through Unidentified Data Accumulated In Cloud
Cloud computing is a rising registering standard in which resources of the processing structure are given as an administration over the Internet. As ensuring as it might be, this standard also conveys quite a few peo...
On Fault Free Response Analysis For BIST
High levels of coverage of classical and nonclassical faults require deterministic test sequences. In this paper, it is suggested that the deterministic test sequences be ordered in such a way that the fault-free outp...
Gender and Name Classification Based on Texture, Local Binary Pattern and Probabilistic Neural Network
In different biometric applications, gender recognition from facial images plays an important role in various applications. This paper investigated Weber’s Local Descriptor for gender recognition which is a texture...
Balancing the Computation-Intensive Function and User Privacy Disclosure at Different Security Levels
We propose a structure for protection safeguarding outsourced utilitarian calculation crosswise over substantial scale numerous encoded areas, which we allude to as POFD. With POFD, a client can get the yield of a ca...