I am Ashay Walke, Final year student of Indian Institute of Technology Kharagpur, India. I am enrolled in the 5 year Integrated Masters course of Mathematics and Computing. I have worked as a Machine Learning Intern at Myntra, India's online fashion market leader, wherein I worked on a Smart AI Based Return Model. I have also worked on Applications of Programming in Modern Algebra during my Summer Research Internship at University of Science, Malysia. I have always been interested in how all operational and strategic decisions can only be made post assimilation and analysis of huge chunks of unsegregated information. I am an active competitive programmer and I love graphic designing.
•Co-authored a paper, "Framework for Automatic Return Approval in a Budget Constrained Environment", under review at CIKM '18 • Built a machine learning model predicting the genuineness of a product return request having auroc 0.81, 0.2 higher than the benchmark • Devised a brand clustering algorithm to reduce the cardiantly (2917) of the feature "brand" with which led in 0.06 auroc increase • Incorporated various models ranging from MLP, XGBoost, CatBoost and lightgbm for SABRE with added feature engineering using fastai
• Simplified the proof of equivalance of Moufang Identities, the known proof was cumbersome and required the knowledge of autotopism • Used SACK in Python, a command-line program for doing elementary computations on small finite groups and printing their cayley tables • Proved that there is no nonassociative loop having order less than 5 with the help of SACK :- Simple Abstract Algebra Calculator in Python
• Wrote article on Drone & Quadcopter analytics which involved codes to find the specific targets in live drone & quadcopter video stream • Extensively involved the use of the openCV library in python for detecting, locating and finding the optimum path for reaching the targets • Leveraged the use of contour properties of different shapes which helped in saving the training of complicated machine learning models • Applied the video enhancement techniques to highlight the details which are obscured involving changing brightness and color contrast
• Assisted the Inside Sales Manager to identify sales opportunities through prospecting & cold relationship development for the US market • Conducted targeted email campaigns for reaching the key decision makers for /hack, a hackathon held at SanFrancisco in August 2016 • Prepared status reports and proposed new business strategies for /hack which witnessed over 900 developers and keynote speakers • Boosted the number of Outbond calls by helping the manager to identify active partners & partners with good business potential
“ It's easier to train a good mathematician the coding parts, than to do it the other way round.” - Sundar Pichai, CEO Google
Statistically Analyzed Sentences from the knowledge base and identified the relation being described using keywords and Assigned confidence scores to country number pairs using gaussian distribution
Developed a Genre recommendation engine in Python using popular collaborative ltering techniques Primarily based on the research done on Single Value Decomposition method during the Netfix Prize competition Implemented and tested other machine learning techniques like Baseline predictor and k-Nearest Neighbor Model
View Project
Used Natural Language Processing to parse information such as Name, Email, Phone Number, skillset and the technology associated with it in any given resume
View Project
Conceptulised, shot and edited ’IIT Kharagpur from AIR’ an aerial view video of theIIT Kharapur campu
View Project