Current Projects

SMAC

SMAC, which stands for State space Modelling And Control theory is an exploratory and implementation project that delves into principles and applications of Probabilistic Modelling using Hidden Markov Models.
This area of study is crucial for a wide range of practical applications, particularly in systems where uncertainty, noise, or incomplete information is inherent.
Hidden Markov Models are powerful techniques used in stochastic analysis that enable us to model of random processes that unfold over time, making it especially well-suited for conducting time series analysis and predicting the evolution of complex systems like the weather, financial markets,
bio-informatics and speech recognition.

SPARC, which stands for Sparse Regression Codes for Erasure Queue Channels is a project that uses concepts of Information Theory, Probability and Stochastic Processes to prove that Sparse Regression Codes can achieve capacity in an Erasure Queue Channel.
An Erasure Queue Channel is a channel model used frequently in Crowdsourcing, Quantum Communications and Multimedia Streaming to capture correlated erasure of transmitted information.
The project aims to use simulations and theoretical proofs to understand such channels better and achieve theoretical capacity during communication.

SUBLIME, which stands for Subdata selection with Big Data and Linear models for Mega dimensions is a project that explores techniques in the intersection of statistics, computational efficiency and big-data analysis.
As datasets grow in both size and complexity, traditional statistical methods often become computationally infeasible or inefficient. We addresses this challenge by focusing on subdata selection algorithms—approaches that extract small, informative subsets from massive datasets while preserving the accuracy of statistical models.
It is particularly based on the implementation of algorithms pertaining to sub data selection in R and C++.
The idea is to introduce the world to this novel method through an accessible, plug and play style library.