I am a highly motivated student pursing Master of Science in Computer Science with specialization in Data Science, at the Erik Jonsson School of Engineering and Computer Science, University of Texas, Dallas. I have experience in C/C++, Java, Python and JavaScript. I enjoy building full-stack applications with CRUD operations. My curious mind compels me to explore the roots of the problem. I am actively learning ReactJS and Node.js through online mediums and open source projects. My strengths include determination, time management and leadership.
I love solving real world problems and that inspires me for competitive programming.
https://leetcode.com/dnyanesht/
https://github.com/DnyaneshT/
-
Projects
MedizinPlus
Feb 2020 – Apr 2020
Project descriptionDeveloped an online pharmacy retailer website with login/registration, searching and checkout modules
through agile development methodology
Incorporated admin module, shopping cart and soft deletion of products
Designed and implemented scalable REST API’s to manipulate data and components throughout the project
Technologies: React, Node.js, MongoDB, Express
Toxic comment classification using BERT embeddings
Feb 2020 – Apr 2020
Project descriptionGiven a dataset of toxic comments, classify the comments into 6 labels; toxic, severe toxic, obscene, threat, insult, identity hate.
Bidirectional Encoder Representations from Transformers (BERT) is a state-of-art NLP pretrained neural network model. The bidirectional encoder generates fixed length vector encoding of the sentence after traversing the sentence in both the directions.
Made use of BERT embeddings to better understand a given comment.
Achieved an accuracy of 89.99% with a hamming loss of 1.84% over 10,000 epochs