College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Computer science & Sliding mode control. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

DeepNet: An Efficient Neural Network for Fake News Detection using News-User Engagements

Abstract: The rise of social media allows every user to share and immediately publish their views. Today, the problem of fake news has obtained significant consideration among researchers due to its harmful nature to deceive the people of the society. It has created an alarming situation in the world. News ecosystem evolved from a small set of trusted and regulated sources to numerous online news sources. Fake news has an adverse impact on society as it may manipulate public opinions. Thus, it is essential to investigate the credibility of news articles shared on social media outlets. In this paper, we have designed an effective deep neural network that is capable of handling not only the content of the news article but also the user-relationships in the social network. We have designed our proposed approach using tensor factorization method. A tensor expresses the social context of news articles formed by a combination of the news, user, and user-group information. Our proposed method (DeepNet) has validated on a real-world fake news dataset: BuzzFeed and Fakeddit. DeepNet outperforms from existing fake news detection methods by employing deep architecture with different kernelsizes convolutional layers combining news content and social context-based features.

Content based medical image retrieval based on new efficient local neighborhood wavelet feature descriptor

Abstract: This paper presents a new class of local neighborhood based wavelet feature descriptor (LNWFD) for content based medical image retrieval (CBMIR). To retrieve images effectively from large medical databases is backbone of diagnosis. Existing wavelet transform based medical image retrieval methods suffer from high length feature vector with confined retrieval performance. Triplet half-band filter bank (THFB) enhanced the properties of wavelet filters using three kernels. The influence of THFB has employed in the proposed method. First, triplet half-band filter bank (THFB) is used for single level wavelet decomposition to obtain four sub-bands. Next, the relationship among wavelet coefficients is exploited at each sub-band using 3 × 3 neighborhood window to form LNWFD pattern. The novelty of the proposed descriptor lies in exploring relation between wavelet transform values of pixels rather than intensity values which gives more detail local information in wavelet sub-bands. Thus, proposed feature descriptor is robust against illumination. Manhattan distance is used to compute similarity between query feature vector and feature vector of database. The proposed method is tested for medical image retrieval using OASIS-MRI, NEMA-CT, and Emphysema-CT databases. The average retrieval precisions achieved are 71.45%, 99.51% of OASIS-MRI and NEMA-CT databases for top ten matches considered respectively and 55.51% of Emphysema-CT database for top 50 matches. The superiority in terms of performance of the proposed method is confirmed by the experimental results over the well-known existing descriptors.

Performance analysis of shunt active power filter using sliding mode control strategies

Abstract: This paper presents a three-phase four-wire shunt active power filter design using two different control strategies. The control techniques analyzed and compared are: sliding mode control (SMC) and proportional-integral control (PI). Current harmonic compensation is achieved by implementation of a Instantaneous Reactive Power Theory for three phase system. The simulation results with and without the shunt active power filter in the system are presented and analyzed. The simulation results show that the SMC controller has a better performance than PI, allowing compensation of reactive power, neutral current, load unbalance and reducing the harmonic level below the limit specified in IEEE-519 standard

Forensic analysis of Malin landslide in India

Abstract: A devastating landslide occurred on 30th July 2014, resulting in the burial of a village of about 40 houses called Malin, in western India and also led to about 160 deaths. The landslide was triggered by heavy rainfall in the area and mass movement of debris. The paper investigates slope failure in the Malin area using back analysis and numerical methods. Site investigation was conducted to obtain representative information of the area. Finite difference analyses using FLAC 2D is performed for the failed slope to determine the possible cause of failure. Analysis results show that slope failure occurred due to the loss of suction strength at the interface between rock and local soil.