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.

Machine Learning approach to classify and predict different Osteosarcoma types

Abstract: Family physicians rarely see a malignant bone cancer because it is hard to find, and most of the time, bone cancer is benign. It is very time-consuming and complicated for the pathologist to classify Osteosarcoma histopathological images. Typically Osteosarcoma classifies into viable, Non-viable, and Non-tumor classes, but intra-class variation and inter-class similarity are complex tasks. This paper used the Random Forest(RF) machine learning algorithm, which efficiently and accurately classifies Osteosarcoma into Viable, Non-viable, and Non-tumor classes. The Random Forest method gives a classification accuracy of 92.40%, a sensitivity of 85.44%, and specificity 93.38% with AUC=0.95.

Investigation on the effect of Cu-doping to ZnTe layers by low-cost electrochemical approach

Abstract: We have reported the electrochemical synthesis zinc telluride (ZnTe) and copper-doped ZnTe (Cu–ZnTe) thin films by electrodeposition technique from non-aqueous electrolyte. Voltammetric investigations under dark and illumination on the deposition were performed to optimize the growth potentials for ZnTe and Cu–ZnTe. The effect of Cu-doping on the structural, morphological, compositional and optical properties has been studied by means of X-ray diffraction, scanning electron microscopy, atomic absorption spectroscopy and UV–Vis spectroscopy. The growth of semiconducting phase was revealed by photovoltammetry. The direction of photocurrent remains unchanged upon the doping of CuCl2 in ZnTe revealed no change in majority charge carriers. The values of the energy band gap of ZnTe and Cu–ZnTe, 10−4 and 2 × 10−4 Cu contents were respectively determined as 2.26, 2.23 and 2.21 eV with sharp absorption edge and enhancement in absorption for the sample grown with 1 × 10−4 M CuCl2. The structure and surface morphology of ZnTe is sensitive to the choice of the deposition potential. The un-doped film deposited at −0.75 and −0.95 V yielded a mixed cubic as well as hexagonal phase of ZnTe along with metallic tellurium. The surface morphology changes from dendritic to globular. Presence of Cu2+ ions in the ZnTe bath not only helped in preventing the formation of tellurium dendrites, it also yielded single hexagonal phase of ZnTe. In case of Cu-doped layer, the surface morphology changes from flake like structure to globular cluster upon changing the growth potential. The un-doped ZnTe samples were Te-rich, whereas Zn/Te ratio for Cu–ZnTe with 10−4 M CuCl2 was found to be nearly unity. The direction of thermoemf was indicative of a p-type response. The increased electrical conductivity in Cu–ZnTe film is associated to increase in the hole density as Cu plays a role of acceptor in ZnTe.

Priority Based Sentiment Analysis for Quick Response to Citizen Complaints

Abstract: Today metropolitan citizens need a common platform to register their complaints. In the traditional telephonic system task of complaint registration is very time-consuming process so, citizen have to wait until call is received by the service executive. The proposed framework is developed for Pune Municipal corporation (PMC) will be very helpful for registering queries in natural language and get immediate response. Understanding the short text is the main challenge of the system as short text do not follow syntax of the written language, short text does not have sufficient statistics to support approach of text mining, short text is noisy and ambiguous. So, traditional Part-of-Speech (POS) tagging tools cannot be easily applied. In proposed framework for understanding natural language semantic knowledge provided by well-known knowledgebase WordNet is used. In prequery citizens inserts complaint to system and get immediate response to query with the help of knowledgebase and machine learning algorithm. In postquery system analyses the citizen sentiment to handle grievance level and accordingly prioritize the citizens by sentiment analysis. The proposed framework will help many organizations to ensure quality service provision and customer satisfaction with less human efforts.

Protection and fault identification in presence of power swing blocking/unblocking function

Abstract: Distance relaying is widely preferred for transmission line protection all over the world. However, the distance relays are badly affected by both, unstable and stable power swing. The occurrence of fault under blocking condition is a severe issue for protection engineer. A new approach for protection of transmission lines when distance relays are blocked is presented using a relatively new signal processing tool known as S transform. The energy content in the current signal is found using the S matrix. To ensure the reliability of the scheme and for the identification of fault types, it is coupled with detection of dc offset and sequence components. The reliability of the scheme is verified by simulating various fault scenarios.

An analysis on data Accountability and Security in cloud

Abstract: Cloud computing is one of the best emerging parameter in IT industry which provides on demand services It uses Central remote server and internet for maintaining data and applications Users can access their files and applications through internet access As it is publically available so data security must be required In this paper, we have analyzed different security and application aspects of cloud computing such as integrity, confidentiality, transparency, accountability, availability, assurance There are many security issues related to cloud computing which includes users data handled remotely by many users, user should know how their data get processed on to the cloud To deal with it many security and accountability models exit Here we discussed some of them like Privacy manager in cloud, RSA based storage security system, Cloud computing data security model and Cloud information accountability framework We have analyzed these four models and finally conclude that which model is suited for which type of application or service Proposed system focused on developing Lightweight Framework for Accountability and Security