Jayant Parashar

Intro

I am from a small city in India named Kaithal. I grew up with a public library inside our house made by my father for our neighborhood's general awareness. In one of the encyclopedias, I read about the Universe and fell in love with the idea of Physics. School curicullum could not feed my curiousity, I started to read on my own. I aspired to be a physicist and followed my dream by studying in famous Hans raj College of University of Delhi. However, my interests started to change as I grew tried of inefficient education system lacking research orientation or passion for the subject. From my various tech book readings, I also came to a conclusion that the most optimal way to solve physics would be to solve intelligence first, based on progress shown by Deep Learning. Therefore, I took a rational decision to change my aspired field of research to Artificial Intelligence.

My love for the abstract and unquenchable thirst for knowledge drive me to work on the problem of intelligence. Solving intelligence is beyond doubt the greatest unsolved problem. I wish to contribute in this great mission. I have developed passion in Deep Learning,Data science, Neuroscience and Machine Learning. I strive to do research in building new complex architectures of Artificial Neural Networks and use them to build an educational application that can feed every child's curiosity.

Research Interests

LONG TERM

I aim to build a cognitive assistant whose underlying architecture is a novel combination of Artifical Neural Networks.

SHORT TERM

tbd

CV

Education

• MS in Artificial Intelligence (In Progress) - University of Georgia (August 2018 - Present)
• CGPA: 4.0
• Post Graduate Diploma in Computer Applications (July 2016 - June 2017)
• Percentage : 78.8%
• Top of the class
• Bachelor of Science in Physics (July 2013 - May 2016)
• Percentage : 76.08%
• with Honors

Experience

• Student Research Assitant (February 2019 - May 2019)
• Strauch Lab, UGA College of Pharmacy

• I worked on predicting binding affinities of peptides with MHCII molecules with Artificial Neural Networks using Python, Keras, SVMlib and Pandas.

• Organizer and Founder (January 2018 - June 2018)
• RLA Academy Meetup Group

• I started an experimental Self Organized Learning Environment (SOLE) with well defined MOOCs’ Tracks in Machine Learning. At its peak, regular attendance was 30 students.

Projects

• Exploring Activation Functions for ANNs on Image classification (April 2019 - May 2019)

• We used Keras to implement Convolutional Neural Networks, Deep Neural Networks and Convolutional Autoencoders with RELU, GELU , Swish, SELU and RELU activation functions. We found that on MNIST and F-MNIST datasets, GELU activation function performed the best.

• Solar Irradiance Prediction using PySpark (April 2019 - May 2019)

• We developed a pipeline which implements a variety of machine learning algorithms for predicting solar irradiance, on Spark. The software’s design centers on execution in Google Cloud Platform along with accessing pertinent data from a Google Storage Bucket.

• Celia Segmentation using Unets (February 2019 - March 2019)

• We segmented Celia from 314 videos of Celia using Unets, a CNN architecture. We used python and Keras. We also preprocessed images by calculating optical flow and variance across pixels.

Honors and Awards

• Top Scorer in Post Graduate Diploma at Kurukshetra University.

Contact

• LinkedIn
• Twitter
• Instagram
• GitHub

For direct line of contact, mail me on : Jayant.Parashar@uga.edu

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';