(SoE) School of Engineering

<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

This paper designs a novel fuzzy competition and attitude based bidding strategy (FCA-Bid), in which the final best bid is calculated on the basis of the attitude of the bidders and the competition for the goods in the market. The estimation of attitude is based on the bidding item's attribute assessment, which adapts the fuzzy sets technique to handle uncertainty of the bidding process as well it uses heuristic rules to determine attitude of bidding agents. The bidding strategy also uses and determines competition in the market (based on the two factors i.e. no. of the bidders participating and the total time elapsed for an auction) using Mamdani's Direct Method. Then the final price of the best bid will be determined based on the assessed attitude and the competition in the market using fuzzy reasoning technique. © 2010 Springer-Verlag Berlin Heidelberg.

World University root node

Computer Science and Engineering

Studies in Computational Intelligence

Automated fuzzy bidding strategy using agent's attitude and market competition

In this work we have extended and implemented an ontology based approach for email classification based on user characteristics proposed by Kim et al.(2007). The approach focuses on finding relationships between user interests and their responses to emails. Rules and Ontology are created using the data and metadata of user characteristics, their preferences and responses to emails. Rules and ontology are then used to predict the response of a user to a new email. In Kim et al. (2007) approach, labels to emails were provided manually by a human expert. We have endeavored to remove the human intervention by developing an Automated Email Categorizer to provide label to an email based on its contents. We have also proposed a new term weighing method for emails to incorporate prominence of subject terms. Finally, we have integrated and tested the Ontology Based Classifier in conjunction with Email Categorizer where the former effectively uses the label provided by latter to classify an email based on user preferences.

KEOD 2010 - Proceedings of the International Conference on Knowledge Engineering and Ontology Development

Automatic email classification using user preference ontology

We propose a graph theoretic technique for recognizing actions at a distance by modeling the visual senses associated with human poses. Identifying the intended meaning of poses is a challenging task because of their variability and such variations in poses lead to visual sense ambiguity. Our methodology follows a bag-of-words approach. Here "word" refers to the pose descriptor of the human figure corresponding to a single video frame and a "document" corresponds to the entire video of a particular action. From a large vocabulary of poses we prune out ambiguous poses and extract 'meaningful' [6] poses - for each action type in a supervised fashion - using centrality measure of graph connectivit [16]. The number of 'meaningful' poses per action is determined by setting a bound on the centrality measure. We evaluate our methodology on four standard activity recognition datasets and the results clearly demonstrate the superiority of our approach over the present state-of-the-art. © 2010 ACM.

ACM International Conference Proceeding Series

Human action recognition in video by 'meaningful' poses

The oil and gas industry forms a vital and large part of the economy of any country. It provides crucial support to transport, manufacturing and energy sectors, produces valuable exports and provides huge employment. This industry along with fertilizer plants, petrochemical plants, etc., which handle hazardous chemicals, are potential targets for deliberate actions by terrorists, criminals and disgruntled employees. The process industries face different levels of threats. It is imperative to analyze the entire threat scenario before taking steps to counter it, otherwise each and every threat will have to be treated as most severe, thus resulting in a huge and wasteful expenditure. The Security Risk Factor Table (SRFT) and a Stepped Matrix Procedure (SMP) have been developed to assess the security risk of oil and gas industry as well as the other chemical process industries. While the SRFT deals with the effects of individual threats, the SMP deals with the cascading/domino effects which a lone, low probability event can cause. A case study of a refinery has been performed to show the application of the ideas presented. © 2010 The Institution of Chemical Engineers.

Chemical Engineering

Process Safety and Environmental Protection

New methodologies for security risk assessment of oil and gas industry

We present a detailed study of low-K/Cu structures under short duration pulse regime and establish a microscopic understanding of breakdown behavior under high current stressing. Random and anomalous behavior of the breakdown characteristics observed under very fast pulsing conditions are explained through electro-thermal instability. A model based on extensive experimental study has been developed to show switching behavior of a conducting path (short failure) due to meltdown of copper. The dumping of critical energy can lead to permanent damage and leads to an open failure. Established breakdown model has been critically linked to the material behavior and extrapolated to understand the TDDB behavior of the low K dielectric. © 2010 IEEE.

Electrical Engineering

IEEE International Reliability Physics Symposium Proceedings

Investigating the electro-thermal origin of breakdown in low-K/Cu dielectrics under short duration over stressed pulsed regime

Drain Extended (DE-NMOS) ESD protection devices exhibit low ESD performance due to poor bipolar action. The purpose of this work is to investigate the role of surface bipolar phenomenon in 90nm DE-NMOS, which triggers failure due to bipolar turn-on, and next study the impact of adding source/drain resistors on their ESD robustness.

Proceedings of the International Symposium on Power Semiconductor Devices and ICs

Improving surface bipolar activity in thin gate oxide DE-NMOS - A critical HV I/O protection element for nano-meter scale technologies

Linear Support Vector Machines (SVMs) have been used successfully to classify text documents into set of concepts. With the increasing number of linear SVM formulations and decomposition algorithms publicly available, this paper performs a study on their efficiency and efficacy for text categorization tasks. Eight publicly available implementations are investigated in terms of Break Even Point (BEP), F1 measure, ROC plots, learning speed and sensitivity to penalty parameter, based on the experimental results on two benchmark text corpuses. The results show that out of the eight implementations, SVM lin and Proximal SVM perform better in terms of consistent performance and reduced training time. However being an extremely simple algorithm with training time independent of the penalty parameter and the category for which training is being done, Proximal SVM is appealing. We further investigated fuzzy proximal SVM on both the text corpuses; it showed improved generalization over proximal SVM. © 2010 IEEE.

ICMLC 2010 - The 2nd International Conference on Machine Learning and Computing

An investigation on linear SVM and its variants for text categorization

This paper presents a linear SVM (Support Vector Machine) Pyramidal Tree (SVMPT) for binary classification tasks. SVMPT is a modified version of SVM based Tree Type Neural Networks (SVMTNN), reported earlier in the literature [1]. Both the algorithms use parameter-less SVM proposed by Mangasarian [2] for learning in each node. While SVMTNN insists on 100 percent training accuracy, linear SVMPT uses predetermined threshold value to determine when to stop adding new nodes. Experimental results on standard binary datasets show that the algorithm has good generalization capability, comparable to linear SVMs. We also present experimental results on extensions of linear SVMPT to multiclass datasets and datasets with fuzzy membership for each datapoint. © 2010 IEEE.

DSDE 2010 - International Conference on Data Storage and Data Engineering

Binary classification using linear SVM pyramidal tree

We present an improved version of One-Against-All (OAA) method for multiclass SVM classification based on a decision tree approach. The proposed decision tree based OAA (DT-OAA) is aimed at increasing the classification speed of OAA by using posterior probability estimates of binary SVM outputs. DT-OAA decreases the average number of binary SVM tests required in testing phase to a greater extent when compared to OAA and other multiclass SVM methods. For a balanced multiclass dataset with K classes, under best situation, DT-OAA requires only (K + 1)/2 binary tests on an average as opposed to K binary tests in OAA; however, on imbalanced multiclass datasets we observed DT-OAA to be much faster with proper selection of order in which the binary SVMs are arranged in the decision tree. Computational comparisons on publicly available datasets indicate that the proposed method can achieve almost the same classification accuracy as that of OAA, but is much faster in decision making. © 2010 Springer Science+Business Media, LLC.

Neural Processing Letters

Fast multiclass SVM classification using decision tree based one-against-all method

The dynamic behavior of liquids in moving containers is a fascinating subject that has attracted the attention of geophysicists, seismologists, engineers, mathematicians and other scientific workers for many years. One of the design issues of the liquid-tank system is the sloshing phenomenon where internal fluid may result in severe damage to the containment system. The slosh dynamics of liquid storage tanks can be strongly influenced by the presence of internal submerged components. The internal components can modify the dynamic characteristics of the liquid tank systems to an advantage. The use of baffles as anti-slosh device is a quite well known example. The baffles cause a variation in the fundamental frequency of liquid oscillations by effectively changing the tank geometry. This paper describes the numerical assessment of lateral sloshing forces developed within liquid filled tanks with various internal components. The main contribution in this paper is that how the slosh dynamics of liquid filled container gets affected due to presence of multiple submerged block and baffle over a partition wall which is still dearth in open literature. © 2010 Begell House, Inc.