VIKASH KUMAR SAH

In anticipation of opportunities that will allow me to utilize my problem-solving skills and to further develop my abilities in the field of Computer Science.

Recommendation System: Expert Identification, (January’19 –December’19), B.Tech Project

• Used Content-Based and Collaborative Filtering to predict experts (researchers) for a user’s query paper.
• Developed a prediction model which considered relevance degree (between experts’ all papers and the query paper), connectivity index (connection b/w similar researchers) and productivity of researchers.
• Used LDA Topic Modeling for finding relevance degree.
• Used A-Miner dataset consisting of details of about 20 lakh papers and 17 lakh researchers.
• Exposure: Recommender System, Information Retrieval, NLP, Topic Modelling

E-Commerce Website, (August’18 – Novemberl’18), Course Project

• Developed an E-Commerce website in Spring framework.
• Prepared Requirement followed by E-R Diagram to do proper analysis of database.
• Used Spring security for authentication & authorization of user SignIn & SignUp.
• Used MySQL 8 Database and normalized data to 2nd Normal Form(2NF).
• Successfully deployed the website for a bounded free period of 15 days.
• Exposure: JAVA, DBMS, Spring, MySQL, CSS, JavaScript.

Object Localization using Deep Reinforcement Learning, (January’18 – April’18), Exploratory Project

• Implemented a model to localize objects in a scene by deforming a bounding box using simple transformation in top-down fashion which are selected by Deep Reinforcement Learning.
• The model was trained and tested on PASCAL VOC 2007 dataset.
• We used pre-trained CNN model to find feature vector of given region of image.
• Exposure: Convolutional Neural Network, Deep Reinforcement Learning.

Student-Teacher Portal cum Forum (Web Portal), (August’17 – November’17), Course Project

• Developed a simple and robust social media website on Django framework.
• Implemented distinct unique profile for every user and middleware were integrated to avoid trespassing.
• A user can login, signup, share post, do comments on a post, post feedback, maintain a friend list, login using social media like Facebook and Github, update their profile and password.
• Exposure: Django, Python, MVC Architecture, Jinja Template Engine.

Aadhar Management System, (January’17 – April’17), Course Project

• Used Binary Search Tree and Hashing Data Structure for storing Person’s details in separate files automatically while entering data and searching using various attributes such as blood group, occupation, gender, etc. simultaneously and give specific results.
• We can perform insertion, deletion, searching(By first name, last name, Blood Group, City).
• Exposure: BST, Hashing, Advantages of Structure in C, File Handling.

ACHIEVEMENTS:

• Council of IITs: Secured All India Rank 2280 in JEE Advanced 2016.
• Kishore Vaigyanik Protsahan Yojana (KVPY) Fellowship Award: Secured All India Rank 720 out of 2 Lakhs students in KVPY 2015 conducted by Indian Institute of Science(IIsc.).