Research Collaboration
Peter the Great St. Petersburg Polytechnic University, Russia
Summer '23
Project Title: Plexus Search – A Search Enumeration
Supervisor: Dr. Elena V. Korchagina
Funding: Peter the Great Saint Petersburg Polytechnic University
Objectives:
The primary objective of the project was the creation of a search algorithm known as "Plexus Search."
This algorithm represents a significant innovation in the field of data retrieval and manipulation within complex data clusters.
The core innovation of Plexus Search lies in its ability to rethink the way data is sought and retrieved within data clusters, particularly in both 1-D and 2-D dimensions.
Research Collaboration
Rzeszow University of Technology, Poland
Summer '23
Project Title: The Silicon Valley Bank Failure: Application of Benford's Law to Spot Abnormalities and Risks
Supervisor: Dr. Grzegorz Zimon
Funding: Rzeszow University of Technology Poland
Objectives:
Recently, in March 2023, Silicon Valley Bank collapsed following unrest prompted by increasing rates. Silicon Valley Bank ran out of money as entrepreneurial investors pulled in-vestments to maintain their businesses afloat in a frigid backdrop for IPOs and individual financing.
The bank's collapse was the biggest since the financial meltdown of 2008 and the second-largest commercial catastrophe in American history.
By confirming the "Silicon Valley Bank" Stock Price Data, further delving into the actual condition of whether there has been any data morphing in the data put forward by the Silicon Valley Bank.
Research Internship
Indian Space Research Organization, India
Autumn '23
Project Title: Advanced Machine Learning-based Spectral Unmixing and Super-resolution of Imaging Infra-Red Spectrometer (IIRS) Data: Applications to surface mineralogy and volatile mapping and implications for petrogenesis.
Supervisor: Dr. Arun P.V.
Funding: Indian Space Research Organization
Objectives:
To develop novel spectral unmixing approaches to map the compositional variability and geological context of the lunar surface.
To develop novel spatial and spectral super-resolution approaches to consider the specific characteristics of the material spectra and the sensor.
To investigate the effectiveness of data augmentations for generating reference spectra samples for the Lunar surface considering the sensor and material characteristics using limited data. 4. To map surface minerals and volatile content and develop new insights into the petrogenesis of the rocks in the study area.
Summer Internship
Sorbonne Université, Paris
Summer '23
Project Title: Physics-informed Neural Networks for forecasting of extreme events
Supervisor: Dr. Tanujit Chakraborty
Funding: Sorbonne University Abu Dhabi
Objectives:
Propose a class of physics-informed differential learning approach for modeling nonlinear dynamical systems, exploiting the physical differentials as the physics-derived loss function.
Proposal is efficient in handling chaotic events in earth science, epidemiology, astrophysics, etc.
Research Internship
National Institute of Advanced Studies, IISc, India
2022 - 2023
Project Title: Physics-informed Neural Networks for forecasting of extreme events
Supervisor: Dr. Nithin Nagaraj
Funding: N/A
Objectives:
ChaosNet, an Artificial Neural Network, built upon the uni-dimensional Luroth Series is used to predict the Star Types.
ChaosNet is conjugated with standard Machine Learners like Support Vector Machine, k - Nearest Neighbours, Decision Tree, etc, for Categorization.
Research Internship
University of Maryland, USA
Winter '23
Project Title: Minimum Average Case Time Complexity for Sorting Algorithms
Supervisor: Dr. Rakesh K. Sharma
Funding: N/A
Objectives:
Sorting Algorithms have been proposed herewith, with various Searching Techniques as intermediate.
The Computational Complexity of the Sorting Algorithm amalgamated with Interpolation Search as an Intermediate Step is compared with Sorting Algorithms amalgamated with Jump Search, Binary Search as an intermediate.
Research Collaboration
Universiti Kebangsaan Malaysia, Malaysia
Summer '23
Project Title: Predicting Cryptocurrency Fraud Using ChaosNet: The Ethereum Manifestation
Supervisor: Dr. Asif Raihan
Funding: Universiti Kebangsaan Malaysia
Objectives:
This research project aims to identify deception and probable fraud in Ethereum transactional processes by utilizing ChaosNet, an Artificial Neural Network constructed using Generalized Luröth Series maps.
The study demonstrates the effectiveness of ChaosNet, a synthetic neural network leveraging chaotic GLS neurons' properties, in performing classification tasks with cutting-edge performance and requiring fewer training samples.
Integration of ChaosNet with established machine learning algorithms enhances its capabilities, surpassing generic results, and providing improved outcomes for identifying deception and fraud in Ethereum transactions.