Hi! I’m Shivanjan

Your friendly neighbourhood Software Engineer

Find out more

Skills

Click the images below to find out more

Python

Pytorch

Tensorflow

NumPy

spaCy

LangChain

OpenCV

HuggingFace

ChatGPT

Go

Kubernetes

C++

C

Experience

Member of Technical Staff 3

Pure Storage

October 2022 - Present

Backend Engineer

Augnito.ai

January 2022 - September 2022

Developer

CloudCover

October 2020 - January 2022

Software Engineer

Oostelijke Onderneming LLC

November 2019 - October 2020

Project Trainee

Maharashtra Knowledge Corporation Limited

July 2018 - August 2019

Student Developer

Google Summer of Code

2018

Education

Biju Patnaik University of Technology

B.Tech Computer Science Engineering

2014 - 2018

Publications

Research Image

Prediction of Human Behavior with TOPSIS

The present paper proposes a new application of the prediction of human behavior using TOPSISas an appropriate tool for data optimization. Our hypothesis was that the analysis of thecandidates with this method was influenced by the change of their behavior. We found that thebehavior change could occur in more than one time span when the behavior of two candidateschanged simultaneously. One of the advantages of this study is that the pattern of the behaviorchange with time is predicted with this method. Another advantage is that the modifications inthe TOPSIS algorithm has made the predictions independent from the need of changing thefuzzy membership degrees of the candidates. This is the first time that these modifications inthis technique with a new application including the numerical analysis of cognitive data arereported. Our results can be used in cognitive science, experimental psychology, cognitiveinformatics and artificial intelligence.

PDF Code


Research Image

Optimization of Machine Learning Algorithms for Proteomic Analysis Using TOPSIS

The present study focuses on a new application of the TOPSIS method for the optimization ofmachine learning algorithms, supervised neural networks (SNN), the quick classifier (QC), andgenetic algorithm (GA) for proteomic analysis. The main hypotheses are that the change in the weights of alternatives could aect the ranking of algorithms. The obtained data confirmed thishypothesis for their ranking. Moreover, adding labor as a cost criterion to the list of criteria did not affect this ranking. This was because candidate 3 had better fuzzy membership degrees thanthe two other candidates concerning their criteria. This work showed the importance of thevalue of the fuzzy membership degrees of the cost criterion of the algorithms in their ranks. The values of the fuzzy membership degrees of the algorithms used for proteomic analysis coulddetermine their priority according to their score differences. One of the advantages of this studywas that the studied methods could be compared according to their characteristics. Another advantage was that the obtained results could be related to the new ones after improving these methods. The results of this work could be applied in engineering, where the analysis of proteins would be performed with these methods.

PDF Code