Post Job Free

Resume

Sign in

Data Analyst Assistant

Location:
La Jolla, CA
Posted:
January 22, 2021

Contact this candidate

Resume:

BALAJI VENKAT SATHIA NARAYANAN

+1-619-***-**** adjm07@r.postjobfree.com

https://www.linkedin.com/in/bsathian https://github.com/bsathian TECHNICAL SKILLS

Languages C, C++, Python, Bash

Packages Pytorch, Scikit-Learn, Numpy, Scipy, Matplotlib, Seaborn, Pandas, MATLAB, Excel Tools Git/Github, LaTeX, CUDA

EXPERIENCE

UC San Diego 2018-Present

Graduate Research Assistant

Used C++, Python and Grid computing to look for new physics beyond the standard model in data taken over three years (> 1 petabyte) from the Large Hadron Collider (LHC), CERN. Improved existing limits on masses of new physics particles by a factor of 2.

Developed CUDA based particle tracking software running on state-of-the-art Nvidia Tesla V100 GPUs to speed up particle tracking by 4 times, for next generation of particle colliders. UC San Diego 2017-2019

Machine Learning based Course Projects

Developed an Image captioning neural network system using Convolutional Neural Networks (CNNs) and Long Short Term Memory networks (LSTMs) augmented with attention models on Pytorch, to produce alt-text to images for better accessibility by the visually impaired.

Developed a Convolutional Neural Network (CNN) based chest X-ray disease identi er on Pytorch to improve diagnosis of pneumonia and allied diseases without human intervention

Built a latent factor model based Recommender system using Scikit-learn and Jupyter Notebooks on data collected from the Goodreads website which can recommend books to users based on their previous reading and review activity.

Developed a MIDI Music le generator on Pytorch using Long Short Term Memory networks (LSTM) and Recurrent Neural Networks (RNN).

UC San Diego Spring 2018

Teaching Assistant/Course data analyst

Used Jupyter Notebooks and Pandas to extract the student scantron responses of 2500 multiple choice quizzes produced by the Scantron machine software.

Tracked overall class performance (250 students) and identi ed parts of the syllabus that needed extra attention and focus using analysis of the student quiz response data. Resulted in instructor redesigning the course and spending more time on the focus topics identi ed EDUCATION

University of California, San Diego Sep 2017- 2022 (Expected) PhD in Experimental Particle Physics

Indian Institute of Science, Bangalore, India Aug 2013- Jun 2017 BS in Physics

RELEVANT COURSES

Machine Learning, Neural Networks and Deep Learning, Parallel Computing in Science and Engineer- ing, Applied Statistics, Computational Statistics



Contact this candidate