Taposh Dutta Roy

This paper describes how Dempster-Shafer (D-S) theory can be used as a tool for improving bi-nary classification prediction performance. While this approach to Dempster-Shafer theory can be used with any binary classification algorithm that outputs probabilities for each example, we show how it can be used with logistic regression specifically. We describe how for each example D-S theory mass functions are obtained for every feature, which are then combined with the ex-ample’s logistic… Show more

This paper describes how Dempster-Shafer (D-S) theory can be used as a tool for improving bi-nary classification prediction performance. While this approach to Dempster-Shafer theory can be used with any binary classification algorithm that outputs probabilities for each example, we show how it can be used with logistic regression specifically. We describe how for each example D-S theory mass functions are obtained for every feature, which are then combined with the ex-ample’s logistic probability as the prior mass function to form the posterior mass function of the example. Show less

This article provides a very simple introduction to predictive modeling using Kaggle's bike sharing as an example.

A simplistic non-personalized recommendation system uses associations to find out top items ( purchases, movies etc) to recommend. For example consider you are a supermarket owner and you keep track of all items sold.

Fork: https://github.com/taposh/recommend_console

It’s estimated that customers consult an average of 11 sources before making a purchase and this year, we saw a 42% jump in spending on Black Friday alone. That’s millions of sources consulted and millions of dollars spent. It was clear that we were in for a huge holiday season as marketers focused on owning the earned media funnel.

Information and influence are crucial elements to the modern day decision-making process. With the rise of bloggers, there is a new breed of experts who are both passionate and knowledgeable about specific topics. Marketers have quickly realized that recommendations by these industry experts and influencers, especially in product categories where consumers conduct research before they buy, are key to moving the market.

Interview with Digital Journal's Michael Kerbs
Digital Journal: Devoted to the ongoing convergence of information technology, news, lifestyle and the arts and showing how technology influences private and professional lives. Articles include the business of digital music, film and publishing, interactive gaming, emerging wireless and broadband content and applications, business-to-business e-commerce, the relationship between digital content and intellectual property law and global… Show more

Interview with Digital Journal's Michael Kerbs
Digital Journal: Devoted to the ongoing convergence of information technology, news, lifestyle and the arts and showing how technology influences private and professional lives. Articles include the business of digital music, film and publishing, interactive gaming, emerging wireless and broadband content and applications, business-to-business e-commerce, the relationship between digital content and intellectual property law and global travel.
Show less

This article analyzes power of Social Media both from an individual's prospective - how it changed people's perspective to things and as well as a consultant's point of view.

In this work, an Indium Antimonide (InSb) quantum well transistor is investigated using full-band Monte Carlo simulations. The steady-state characteristic of the device is first analyzed, showing particle transport along the two-dimensional electron gas (2DEG). The small-signal behavior of the device is also investigated. Finally, the noise analysis is performed, allowing for a two-dimensional mapping of the noise within the… Show more

In this work, an Indium Antimonide (InSb) quantum well transistor is investigated using full-band Monte Carlo simulations. The steady-state characteristic of the device is first analyzed, showing particle transport along the two-dimensional electron gas (2DEG). The small-signal behavior of the device is also investigated. Finally, the noise analysis is performed, allowing for a two-dimensional mapping of the noise within the device.

http://taposh.files.wordpress.com/2010/06/poster.pdf

Google Books:
http://books.google.com/books?id=3vATanlMIIsC&lpg=PA358&ots=Gf5xA45KqN&dq=monte%20dutta-roy%20T&lr&pg=PA358#v=onepage&q=monte%20dutta-roy%20T&f=false Show less

Abstract—This paper is a review of the developments in Instruction level parallelism. It takes into account all the changes made in speeding up the execution. The various drawbacks and
dependencies due to pipelining are discussed and various solutions to overcome them are also incorporated. It goes ahead in the last section to explain where is the new research leading us.
Index Terms—Branch Prediction, Exceptions, Instructional level parallelism, pipelining.

Abstract—Division algorithms have been developed to reduce latency and to improve the efficiency of the processors. Floating point division is considered as a high latency operation. This papers looks into one such division algorithm, examines the hardware block diagram and suggests an alternative path which may be cost effective.