About

Ayush Surana is a third-year undergraduate student of the department of Chemical Engineering enrolled in its B.Tech course. He is a Core Organizing Team Head at Kshitij, IIT Kharagpur, where he applies his technical skills to conflux the spheres of technology and innovation. Ayush has outlined specific goals to help him grow holistically, including getting an experience of the corporate world, an exposure in the field of artificial intelligence and machine learning, and developing a network of intellectuals. Ayush is attracted to challenging real-life problem-solving, whether it is academics or management. He has inculcated leadership and management skills through numerous positions of responsibilities that he has undertaken. Ayush has constantly aimed at pursuing diverse interests, and in addition to academic proficiency, he has inculcated in himself leadership and management skills. Ayush has relevant experience of 0.36 years. He has worked as a Senior Member at Kharagpur Data Analytics Group, where he designed a Stock Price Predictor for stocks in the Indian equity market using the LSTM model with an R2 score of 0.925891. He also applied various models such as Logistic Regression for Sentiment Analysis of tweets with a best F1 score of 0.81342. Ayush trained Random Forest Classifier on an imbalanced dataset of Credit Card Fraudulent Detection with an accuracy of 0.99961. Ayush has also worked as a Research Intern at the Institute of Business and Technology, where he developed an Information Retrieval System to get summarized information on customized user input. He designed an automated web extractor for extraction of abstracts through Web of Science using Selenium WebDriver. Ayush deployed transformer-based models such as BERT, ERNIE as transfer learning for text classification and abstract analysis. He tuned the Hyperparameters of models and achieved an accuracy of 91% on multi-class text-based classification. Ayush also implemented a Text-to-Text transfer transformer as a summarizer to give an abstract summary of the input topic. He modeled to reduce the average reading time of abstracts of research paper by 25% when compared to the manual process. Ayush's TechStack includes Sentiment Analysis, Random Forest, ML, Research Scientist, Information Retrieval, BERT, Text Classification, Web, Hyperparameter, and Selenium.