Curriculum Vitae ---------------- Personal Information -------------------- Name : Abhishek Seth DOB : May 28, 1983 Email : abhiseth@it.iitb.ac.in Home page : http://www.it.iitb.ac.in/~abhiseth Academic Record --------------- M.Tech - Information Technology 2004 to 2006 KReSIT, IIT Bombay Advisor: Prof. Bernard Menezes CPI - 9.7/10 B.E. - Information Technology 2000 to 2004 Shri Vaishnav Institute of Technology and Science, Indore (M.P.) Rajiv Gandhi Technical University of Madhya Pradesh Percentage: 79.84 Graduated First class with Hons. Interests --------- 1. Forecasting of Time Series 2. Data Mining Technologies 3. Algorithms Projects -------- 1. MTech Thesis Title: Improve Forecasting Performance of Time Series by Various Post Processing Techniques. Guide: Prof. Bernard Menezes Desc: The purpose of this project is to investigate the advantage of using a large number of forecasting models (called experts) simultaneously to improve the forecasting performance. Various aspects comes into consideration such as the class of experts to use and how to combine their judgements in a useful way. The combining of experts is useful when it decreases the percentage error of the final judgement and/or increases the confidence in it. Our method of forecasting on demand sales series is achieving a significant improvement ( currently 9% ) over the standard Holt Winter Method. We are working on refining our system of forecasting. 2. MTech Seminar Title: Wireless Security Guide: Prof. Bernard Menezes Desc: This seminar was concentrated on various issues in Wireless Security. A significant part of this seminar was consisting of intrusion detection in Wireless Scenerio. 3. BE Project Title: Real Time Processing of Audio/Video Streams Guide: Prof. Sunita Varma Desc: In this project, we have worked on enabling the real time processing viz. encryption/compression in audio/video streams. The complexity comes into picture when we want to maintain the real time performance, which requires strong yet simple encryption techniques. The compression techniques desirable for such service should work fast enough which along with the requirement of low bit rate compression, increases the complexity of this project a step further. We achieved these challenges upto a satisfactory level of performance. 4. Forecast Engine: We developed a forecast Engine which takes the forecast values from various backend systems and combine them to produce more usefull results. Various backend system used in this project are SAS, Matlab, Neural Network programs developed in Java. The architecture of this project is made flexible enough to accomodate new backends with minimal effort. Currently this project find its application in MTech Thesis work of one of my classmate. 5. QOS Agent for DiffServ(Differentiated Service) Network: The aim of this project was to develop an agent which can automatically configure the network nodes of diffserv network to enable Quality of Service(QOS). The QOS requirement of applications is captured in the form of token bucket parameters which are passed to all edge/core routers in the path. Reservation is made only if sufficient resources are available. This project finds its application in practical enterprise networks where Differentiated Service is enabled and cumbersome task of system administrator has to be minimized. Computer Skills --------------- Software: SAS, MATLAB. Application/Web Servers: LDAP,Apache, Programming Languages: C/C++, Java, Scripting Language: shell scripting, Perl, PHP. Web Technologies: HTML,XML Databases: SQL Server, MySQL, and Postgres. Platforms: Linux, Windows References ---------- 1. Prof. Bernard Menezes Professor, KReSIT, IIT Bombay Declaration ------------ I hereby declare that the above mentioned information is correct to best of my knowledge. Abhishek Seth IIT Bombay Feb, 2010.