Anshik Bansal

ABOUT ME

Passionate to learn and develop machine learning systems, built upon strong fundamentals and designed to scale. Working on solving challengi

Passionate to learn and develop machine learning systems, built upon strong fundamentals and designed to scale. Working on solving challenging problems in unstructured data using NLP, Deep learning and PGMs

Hi, As a Machine learning expert i am ridiculously committed to your learning and growth in this fast evolving field. I have a total experience of 3+ years in Machine Learning. I keep myself up-to-date with latest development and constantly think where improvements can be made in existing pipelines. I put deep emphasis on having a well-maintained code and complete explainations which brings clarity for my students and do not hesistate from spending a little extra time if the concept is not understood well.

English

Chennai (+05:30)

Joined January 2019

EXPERTISE

Machine Learning

3 years experience

NLP (Natural Language Processing)

4 years experience

Python

4 years experience

SOCIAL PRESENCE

GitHub

bayesian_machine_learning

My attempt to understand the bayes world

Jupyter Notebook

node-web-server

Node test app

JavaScript

Community Posts

Natural Language Processing, oh I lost track - Part 1

A succinct attempt to understand Natural Language Understanding one post at a time.

EMPLOYMENTS

Data Scientist

ZS Associates

2017-07-01-Present

https://www.linkedin.com/in/anshik-bansal-8b159173/

Python

Computer Vision

NLP (Natural Language Processing)

Python

Computer Vision

NLP (Natural Language Processing)

TensorFlow

Keras

PROJECTS

Hybrid Approach for End-to-End Entity Recognition and LinkingView Project

2018

Dynamic entity recognition and linking entities with relationship is one of the most challenging problems in Biomedical NLP. A version ...

Dynamic entity recognition and linking entities with relationship is one of the most challenging problems in Biomedical NLP. A version of this problem was presented as a challenge as part of the BioCreative OHNLP challenge 2018 where we had to identify family members and link them to their respective prevalent diseases using unstructured textual notes of family history (1) We leveraged open-source meSH ontology, coupled with word2vec trained on textual data to train a domain-specific word vector(esp. for diseases) and then segmented into relevant clusters of similar semantic information using K-means

Python

Machine Learning

NLP (Natural Language Processing)

Python

Machine Learning

NLP (Natural Language Processing)