Having been selected in prestigious [email protected] 2017 program among thousands of students worldwide, I worked on a project titled - "Supervised Q-walk for Learning Vector Representation of Nodes in Networks" in the realm of network analysis which involved developing a method for learning vector representation of nodes in a network. The method was inspired by a recently proposed node2vec framework. The latter framework is unsupervised, whereas my approach was a supervised adaptation leveraging label information for improving upon the quality of the learned representation of nodes. The project was developed using Python 3.5, NetworkX, Numpy, Matplotlib, Scikit-learn, Gensim on Intel Xeon powered Ubuntu 16.04.
This project was done individually as part of [email protected]EPFL 2017 programme.
"...Throughout his internship, Naimish put an extraordinary amount of passion and hard work
into his project. He ran extensive experiments and regularly provided detailed updates to
me. Finally, he compiled his findings in a thorough report. In the process, Naimish has
broadened his machine learning expertise significantly through self-directed study.
I am confident Naimish will fare very well in the remainder of his Master’s degree and wish
him the very best for his future career. "
- Prof Robert West, Assistant Professor, Data Science Lab, School of Computer and Communication Sciences, EPFL
Letter of Appraisal can be seen at http://bit.ly/2wivAON
The pre-print of the paper is available at: https://arxiv.org/abs/1710.00978. It has been reviewed and accepted for presentation at the IEEE Technically Sponsored Intelligent Systems Conference (IntelliSys) 2018, to be held from 6-7 September 2018 in London, United Kingdom.