Shubhankar Panse

** ****** ** #*, ****** City, NJ 07306 201-***-**** ac308n@r.postjobfree.com LinkedIn: https://goo.gl/9Ebctn Github: https://github.com/ssp573 Education

New York University, New York, USA May 2019

Courant Institute of Mathematical Sciences

Master of Science, Computer Science (GPA – 3.67/4) Relevant courses- Foundations of Machine Learning, Fundamental Algorithms, Programming Languages Savitribai Phule Pune University, Pune, India Jul 2017 Pune Institute of Computer Technology

Bachelor of Engineering, Computer Engineering (GPA – 3.65/4) Relevant courses- Theory of Computation, Data Mining, Cloud Computing, Database Management Skills

Programming/ Scripting Languages: Java, C/ C++, Python, Scala, Lua, R, Shell, Perl, Scheme, Prolog, ML, Javascript, .NET

Operating Systems: Windows, Ubuntu, Fedora

Database Development: SQL, MongoDB

Other Tools: NLTK, spaCy, Torch, Android Studio, Pandas, TensorFlow, Tableau Professional Experience

Right Steps Consultancy, Intern (Python, NLTK, spaCy) Jan 2017 - Jun 2017

• Cleaned databases to find potentially similar profiles and formulated a similarity score using Levenshtein Distance, Soundex algorithm in Python.

• Summarized lengthy legal contracts by extracting only the important information using NLTK, spaCy and algorithms like RAKE, Textrank in Python.

CVRanker, Intern (SQL) Aug 2016 - Nov 2016

• Designed an SQL database to hold customer data and migrated data stored in MS Excel.

• Maintained a digital presence by using various tools like MailChimp and Hootsuite. Academic Projects

Predicting the impact of Movie Dialogues - A Class Imbalance problem (Python, Sklearn, Adaboost)

• Classified movie dialogues into impactful and non-impactful classes. This is a class imbalance learning problem where there are disproportionately high samples of the non-impactful type which hinders the overall performance of the classifier.

• Solved the class imbalance problem using various ways like assigning more prominent weights to minority class than the majority one, Agglomerative clustering, oversampling and undersampling along with variants of the Adaboost algorithm.

Artistic Style Transfer Using Deep Learning (Torch, Lua, CNN)

• Regenerated real world images in an artistic style using a linear combination of style and content features extracted from a pre-trained Convolutional Neural Network VGGNet.

• Developed a good understanding of Convolutional Neural Networks along with it’s implementation.

• Worked with ‘Thumbworks Technologies Pvt. Ltd.’ for this project. Book My Turf (Android Studio, Java, SQL)

• Built an Android Application to facilitate reserving sports amenities.

• Designed an SQL database to hold the data. Used Android studio for the app development and javascript to link the them.

Leadership and Awards

• Secretary, PICT IEEE Student Branch (PISB) Jun 2015 - Jun 2017

• Darrel Chong Student Activity Award – PICT IEEE Student Branch Mar 2016

Contact this candidate