Shama F. Barna, PhD

Data Science Process Development

217-***-**** Permanent Resident

adcl7p@r.postjobfree.com Linkedin Profile GitHub Profile EDUCATION:

University of Illinois at Urbana Champaign 2012 - 2018 PhD in Mechanical Science and Engineering, GPA (3.55/4) University of Illinois at Urbana Champaign 2009 - 2011 MS in Mechanical Science and Engineering, GPA (3.55/4) Islamic University of Technology, Bangladesh 2004 - 2007 BS in Mechanical Engineering, GPA (4.91/5)

SKILLS:

Coding Skills: Python, SQL, Matlab, Jupyter Notebook Data Science Skills: pandas, numpy, sklearn, nltk, genism, keras etc. Machine Learning Skills: Supervised/Unsupervised algorithms, Neural Network, Feature Engineering etc. Statistical Skills: Inferential Statistics, Hypothesis Testing, Bayesian Statistics etc. Other Skills: PyTorch, Scala, PySpark

EXPERIENCE:

Process Engineer and Data Scientist at Intel Corporation 4/2018 - Present

● Act as a Process Engineer and Data Scientist for the development of next-generation micro-processors

● Worked with a front-line data analysis team to investigate root causes of tool downtime

● Performed data aggregation, analysis and visualization to improve operational efficiency by 15 hours/week

● Work with vendors and cross-functional teams to troubleshoot problems in the process flow

● Identify and fix hardware problems related to wafer patterning tools for high volume production PROJECTS:

1. Multiclass Text Classification for Identifying TV Show Characters (Link to Project Report) 2019

● Web-scraped data from transcripts available online for the TV show “Friends”

● Implemented interactive visualizations to find insights from the transcripts of the TV series

● Explored different word embedding techniques for Machine Learning modeling

● Developed a baseline model with dummy classifier to benchmark performance of other classifiers

● Evaluated prediction capabilities of linear classifiers using cross-validation and hyper-parameter tuning

● Utilized oversampling and undersampling techniques to address imbalance in the dataset

● Investigated how modeling performance varies for increasing numbers of classes in the dataset

● Developed Deep Learning model for better prediction capabilities 2. Collaborative Filtering Based Book Recommendation Engine (Link to Project Report) 2019

● Utilized exploratory data analysis to identify users’ reading behaviors and trends in the book market

● Developed collaborative filtering Machine Learning models to predict user ratings

● Demonstrated improved modeling performance with hyper-parameter tuning

● Implemented non-personalized and personalized recommendation system for new and current users

● Developed a smart filtering system for book search utilizing data wrangling CERTIFICATION:

● Springboard Academy 7/2019 - Present

Data Science Career Track Bootcamp

Contact this candidate