Vishwakarma Institute of Information Technology

Pune, India

Vishwakarma Institute of Information Technology

Pune, India
Time filter
Source Type

Malji P.,Vishwakarma Institute of Information Technology | Sakhare S.,Vishwakarma Institute of Information Technology
Proceedings of 2017 11th International Conference on Intelligent Systems and Control, ISCO 2017 | Year: 2017

Feature selection is an essential method in which we identify a subset of most useful ones from the original set of features. On comparing results with original set and identified subset, we observe that the results are compatible. The feature selection algorithm is evaluated based on the components of efficiency and effectiveness, where the time required and the optimality of the subset of the feature is considered. Based on this, we are modifying the fast clustering feature selection algorithm, to check the impact of entropy correlation coefficient on it in this paper. In the algorithm, the correlation between the features is calculated using entropy correlation coefficient instead of symmetric uncertainty and then they are divided into clusters using clustering methods based on the graph. Then, the representative features i.e. those who are strongly related to the target class are selected from them. For ensuring the algorithm's efficiency, we have adopted the Kruskal minimum spanning tree (MST) clustering method. We have compared our proposed algorithm with FAST clustering feature selection algorithm on well-known classifier namely the probability-based Naive Bayes Classifier before and after feature selection. The results, on two publicly available real-world high dimensional text data, demonstrate that our proposed algorithm produces smaller and optimal features subset and also improves classifiers performance. The processing time required for the algorithm is far less than that of the FAST clustering algorithm. © 2017 IEEE.

Sathe J.B.,Vishwakarma Institute of Information Technology | Mali M.P.,Vishwakarma Institute of Information Technology
Proceedings of 2017 11th International Conference on Intelligent Systems and Control, ISCO 2017 | Year: 2017

Neural Network(NN) and fuzzy systems are suitable for determining the input-output relationships. NN contend with numeric and quantitative information whereas fuzzy systems can handle symbolic and qualitative information. Coupling of Neural Network and Fuzzy Logic results in an intelligent crossbreed system widely referred to as Neuro-fuzzy system (NFS) that exploits the most effective qualities of these two approaches expeditiously. The coupled system combines the human alike logical reasoning of fuzzy systems with the training and connectedness structure of neural network. In this paper, we propose a method for performing Sentiment Classification using an NN and fuzzy set theory. In this method input reviews are fuzzified by using Gaussian membership function and fuzzification matrix is build. This matrix is transposed and passed to Multilayer Perceptron Backpropagation Network(MLPBPN). © 2017 IEEE.

Nitsure S.P.,Vishwakarma Institute of Information Technology | Londhe S.N.,Vishwakarma Institute of Information Technology | Khare K.C.,NBN Sinhgad Technical Institute Campus
Ocean Engineering | Year: 2012

Forecasting of wave heights is essential for planning and operation of maritime activities. Traditionally, wave heights have been predicted using physics-based models, which rely primarily on the energy balance equation. More recently, soft computing techniques such as Artificial Neural Network (ANN), Genetic Programming (GP) have been used to generate forecasts with leads times from a few hours to several days. However, the forecast accuracy of both methods could be improved, particularly at peak wave heights, and at higher lead times. This paper forecasts the wave heights with lead times of 12 h and 24 h using GP. The data are obtained from two locations, along the North American and Indian coastlines. Wind information is used as an input. The modeling procedure relies heavily on the parameter kurtosis, or fourth moment. The forecasts are satisfactory, especially for the peak wave heights formed by the extreme events like hurricanes. © 2012 Elsevier Ltd.

Londhe S.,Vishwakarma Institute of Information Technology | Charhate S.,Datta Meghe College of Engineering
Hydrological Sciences Journal | Year: 2010

Accurate forecasting of streamflow is essential for the efficient operation of water resources systems. The streamflow process is complex and highly nonlinear. Therefore, researchers try to devise alterative techniques to forecast streamflow with relative ease and reasonable accuracy, although traditional deterministic and conceptual models are available. The present work uses three data-driven techniques, namely artificial neural networks (ANN), genetic programming (GP) and model trees (MT) to forecast river flow one day in advance at two stations in the Narmada catchment of India, and the results are compared. All the models performed reasonably well as far as accuracy of prediction is concerned. It was found that the ANN and MT techniques performed almost equally well, but GP performed better than both these techniques, although only marginally in terms of prediction accuracy in normal and extreme events. © 2010 IAHS Press.

Joshi J.,Vishwakarma Institute of Information Technology | Mundada G.,Pune Institute of Computer Technology
Proceedings of the 2010 5th International Conference on Information and Automation for Sustainability, ICIAfS 2010 | Year: 2010

In wireless mobile communication systems, the radio spectrum is limited resource. However, efficient use of such limited spectrum becomes more important when the two, three or more cells in the network become hot-spot. The use of available channels has been shown to improve the system capacity. The role of channel assignment scheme is to allocate channels to cells in such way as to minimize call-blocking probability or call dropping probability and also maximize the quality of service. In this paper attempts are made to reduce call-blocking probability by designing hybrid channel allocation (HCA) which is the combination of fixed channel allocation (FCA) and dynamic channel allocation (DCA). A cell becomes hot-spot when the bandwidth available in that cell is not enough to sustain the users demand and call will be blocked or dropped. A simulation result shows that HCA scheme significantly reduces call-blocking probability in hot-spot scenario and compared with cold-spot cell. This hot-spot notification will request more than one channel be assigned to the requesting cell, proportional to the current hot-spot level of the cell. Furthermore, all channels will be placed in a central pool and on demand will be assigned to the base station. That will be helpful to reduce call-blocking probability when cell becomes hot-spot. When a call using such a borrowed channel terminates, the cell may retain the channel depending upon its current hot-spot level therefore HCA has comparatively much smaller number of reallocations than other schemes. It also shows that it behaves similar to the FCA at high traffic and to the DCA at low traffic loads as it is designed to meet the advantages of both. © 2010 IEEE.

Prasad J.R.,Vishwakarma Institute of Information Technology | Kulkarni U.V.,Shri Guru Gobind Singhji Institute of Engineering and Technology
Proceedings - 3rd International Conference on Emerging Trends in Engineering and Technology, ICETET 2010 | Year: 2010

the field of Handwriting recognition has evolved over the past three or four decades into a broad based activity which has had a measurable impact on applications. Some of the most significant practical impact has occurred in the past decade in handwriting recognition. Successful application of the established methods requires good understanding of their behavior and how well they match a particular context. Difficulties can arise from either the intrinsic complexity of a problem or a mismatch of methods to problems. Many emerging applications of involve complicated high-dimensional pattern spaces, small amounts of data-per-dimension, low signal-to-noise ratio, poorly specified statistical distributions, and anomalous statistical outliers. In some cases these difficulties are compounded by distributed data collection requirements that impose constraints on data integration and decentralized decision making. This creates both challenges and opportunities for Handwriting recognition research. This survey divides various approaches to handwriting recognition in nine different categories. Authors explore resent trends in Handwriting recognition and describe the areas of challenges and some possible solutions. © 2010 IEEE.

Chinchanikar S.,Indian Institute of Technology Kanpur | Choudhury S.K.,Indian Institute of Technology Kanpur | Kulkarni A.P.,Vishwakarma Institute of Information Technology
Advanced Materials Research | Year: 2013

In the present work, effect of work material hardness and cutting parameters on chip-tool interface temperature was investigated during turning of AISI 4340 steel hardened at two different levels of hardness 35 and 45 HRC, respectively, using CVD applied multi-layer TiCN/Al2O3/TiN coated carbide inserts. A tool-work thermocouple principle was used to measure the interface temperature during turning. The correlation coefficient between experimental and predicted values of interface temperature found close to 0.95, which showed that the developed model is reliable and could be used effectively for predicting the interface temperature for the given tool and work material pair and within the domain of the cutting parameters. Experimental observations indicate that the interface temperature is higher for harder work material and get affected mostly by cutting speed followed by feed. However, depth of cut has little influence on interface temperature irrespective of the hardness of the workpiece. © (2013) Trans Tech Publications, Switzerland.

Prasad J.R.,Vishwakarma Institute of Information Technology | Kulkarni U.V.,Shri Guru Gobind Singhji Institute of Engineering and Technology
Proceedings - International Conference on Electronic Systems, Signal Processing, and Computing Technologies, ICESC 2014 | Year: 2014

This paper presents implementation of an Adaptive Neuro Fuzzy Classifier (ANFC) for recognition of isolated handwritten characters of Gujrati based on [2]. Authors aim to compare the performance of ANFC with weighted k-NN classifier proposed in [1] by them. Fuzzy classification is the task of partitioning a feature space into fuzzy classes. Authors exploit the method of employing adaptive networks based on [2] to solve a fuzzy classification problem. System parameters, such as the membership functions defined for each feature and the parameterized t-norms used to combine conjunctive conditions are calibrated with back propagation. Towards this aim, authors use a supervised learning procedure based on Scaled Conjugate Gradient (SCG) algorithm to update parameters in an adaptive network. Next, this architecture is deployed for the character recognition problem. From the experimental results, it is summarized that although adaptively adjusted classifier performs well as far as time complexity is concerned but fails to achieve better recognition rates than weighted k-NN. The results are discussed from the viewpoint of feature extraction methods discussed in [1] and their effectiveness on neuro fuzzy classifiers. © 2014 IEEE.

Kulkarni A.A.,Vishwakarma Institute of Information Technology | Khandekar P.D.,Vishwakarma Institute of Information Technology
IEEE-International Conference on Advances in Engineering, Science and Management, ICAESM-2012 | Year: 2012

Clock distribution network forms an integral part of any digital circuit. It consumes a large part of the total circuit power, which is not desirable. Different techniques are employed up till now to reduce the clock power. In this paper we have demonstrated how clock power can be reduced significantly by distributing it at reduced supply voltage. The clock distribution network is designed and simulated in 0.25m technology. It is simulated at different frequencies of 10MHz, 100MHz, 200MHz, 250 MHz and 400MHz achieving power reduction of about 53%, 44%, 41%, 24% and 5% respectively. © 2012 Pillay Engineering College.

Chinchanikar S.,Vishwakarma Institute of Information Technology | Choudhury S.K.,Indian Institute of Technology Kanpur
International Journal of Machine Tools and Manufacture | Year: 2015

The researchers have worked on many facets of machining of hardened steel using different tool materials and came up with their own recommendations. Researchers have tried to investigate the effects of cutting parameters, tool materials, different coatings and tool geometry on different machinability aspects like, the tool life, surface roughness, cutting forces, chip morphology, residual stresses and the tool-chip interface temperature under dry and/or semi-dry and/or flood cooling environment during machining of hardened steels while many of them have ventured to characterize the wear phenomenon. Good amount of research has been performed on an analytical and/or numerical and/or empirical modeling of the cutting forces, tool-chip interface temperature, and tool wear under orthogonal/oblique cutting conditions during machining of hardened steels. This paper presents a comprehensive literature review on machining of hardened steels using coated tools, studies related to hard turning, different cooling methods and attempts made so far to model machining performance(s) so as to give proper attention to the various researcher works. © 2014 Elsevier Ltd. All rights reserved.

Loading Vishwakarma Institute of Information Technology collaborators
Loading Vishwakarma Institute of Information Technology collaborators