picture of ankit
Ankit Mishra
Ph.D. Candidate, Viterbi School of Engineering
University of Southern California

Bio


I am currently a Ph.D candidate in the Department of Chemical Engineering, I work with Prof Priya Vashishta and Prof Aiichiro Nakano at Collaboratory of Advanced Computing and Simulations (CACS). My research work involves studying materials using Machine Learning techniques and designing high performance simulations to study various materials such as polymers, ceramics etc. My current work is focussed on accelerated design of material using Graph Neural Networks. Prior to this I have worked on designing and implementing research applications for Immersive Visualization of Materials Simulation using Virtual Reality, Active Learning Using Bayesian Optimization to find optimal energy surfaces for materials under extreme conditions, Multi-Objective Genetic Training and Uncertainty Quantification of complex material systems

I am also currently pursuing my Masters in Computer Science. I have taken courses in Object Oriented Programming, Operating Systems, Computer Networks, Web Development, Data Mining, Machine Learning, Maathematical Optimization, Probability and Statistics. I have good experience with various web development tools such React.JS, Node.JS, Express.JS. I also have considerable experience using various machine learning libraries in python such as sklearn, tensorflow, numpy, pandas etc.

Before this, I attended Indian Institute of Technology (IIT), Varanasi where I got my undergraduate degree in Chemical Engineering (in 2014).

I am originally from Varanasi, India. I started my Ph.D. in Fall 2015 and currently live in Los Angeles, California.

Education


University of Southern California
PhD in Chemical Engineering , Fall 2015 - Spring 2021

Pursuing PhD in Chemical Engineering from USC, my research primarily focuses on high throughput simulations for accelerated discovery of materials. My research work also involves active learning based material discovery framework, uncertainty quantification and multi-objective evolutionary algorithm based forcefield fitting.
University of Southern California
Masters in Computer Science , Fall 2019 - Spring 2021

Pursuing masters in computer science from USC with emphasis on developing data intensive applications for commerical and scientific use. I am also experienced at developing web based applications. Please have a look at development projects for more details.
Indian Institute of Technology, Varanasi
Bachelors in Chemical Engineering , August 2010 - May 2014

I completed by Bachelor of Technology in Chemical Engineering from prestigious IIT (BHU).

Research


Accelerated Discovery of Materials Using Graph Neural Networks
Designed a simulation framework to estimate electrical properties of polymers using Molecular Dynamics Simulation. This framework is a high throughput framework capable of handling 1000s of polymer simulations at once. Further, we used Graph Neural Networks to learn various complex features for each polymers to predict the based structures which can give desirable electrical properties.
Ankit Mishra, Pankaj Rajak, Ajinkya Deshmukh, Lihua Chen
Multi University Reseach Initiative (MURI), Summer 2019 - Spring 2020
Uncertainty Quantification of Kevlar Potential Energy Surface Using Bayesian Optimization
The potential energy surface for Kevlar under shock is too complex making it challenging to find a global minima useful for its design and fabrication. Furthermore, computation on each structure is extremely costly. Thus, we employ Bayesian optimization technique to efficiently find the optimal solution.
Subodh Tiwari, Kohei Shimamura, Ankit Mishra, Fuyuki Shimojo
Project, Spring 2019
EZFF A Python based Framework for Uncertainty Quantification and Genetic Training of physical simulation
Used various Evolutionary Algorithms such as Genetic Algorithms, Particle Swarm Optimization, Differential Evolution to quantify the uncertainty involved in multi-objective optimization of general material simulations. Here, we used a famous Multi Objective Optimization library Platypus to build a framework incorporating various popular simulation software used in Materials simulations such as LAMMPS, GULP, VASP, RXMD
Aravind Krishnamoorthy and Ankit Mishra
Project, Spring 2018 - Spring 2019
GEARS-XR: Collaborative Mixed Reality Platform for Scientific Visualization
Visualization plays a critically important role in modern scientific research project. It is a very powerful tool to express results from observation and experiments, and visually provides scientists with an intuitive overview of the results. We used "Acer Windows Mixed Reality Headsets" which has 1440x1440 resolution and 90 fps at maximum, as a head-mounted display to immersively visualize materials simulations in real time. GEARS-XR consist of three components: LAMMPS interface, multi-user component, and the rendering component. Different from any other applications, e.g., games or movies, scientific visualization require strict criteria for preciseness, which makes multi-user component and rendering component non-trivial technical building blocks.
Ankit Mishra, Vaibhav Desai, Rohan Rout and Hikaru Ibayashi
Augmented and Virtual Reality Project, Fall 2018

Development


Burger Builder: A responsive application using ReactJS
This application is written in React.JS and bootstrapped using create-react-app. The application's backend is hosted on Firebase platform. It uses React router and redux library to efficiently route through mutli pages simultaenously storing the user data with appropriate validation.
Ankit Mishra
Project, Summer 2020
Responsive React.JS, Node.JS based news app using Guardian and NYTimes APIs
Developed responsive news app using React.JS frontend and Node.JS backend. The app was designed in various components to enhance reusability and hosted on Amazon EC2 web server. The news information for this application is fetched from Guardian and NYTimes API to get real time news data. The backend utilizes REST calls using Node and Express.Js frameworks. The styling for various components in this app is done using React-bootstrap.
Ankit Mishra and Taufeq Razzakh
Project, Spring 2020
News App using Python Flask, HTML, CSS and JS
A web app designed to display top news from Fox and CNN news using google news api. The app also has a feature to seach for news based on keywords, categories, sources and published date, This app was built using Flask backend and Javascript frontend along with AJAX for asynchronous API calls.
Ankit Mishra and Taufeq Razzakh
Project, Spring 2020
Mining of Massive Datasets, Python, Apache Spark
Implemented following data mining algorithms using Apache Spark Framework in Python:
  • SON Algorithm (parallel and scalable version of Apriori algorithm) to find possible combinations of frequent itemsets in yelp dataset
  • Community detection using Spark GraphFrames library and Girvan-Newman Algorithm
  • Streaming Algorithms on Yelp dataset: Bloom filtering to test existing user membership, Flajolet-Martin algorithm to estimate the number of unique users and fixed sized Reservior sampling algorithm
  • BFR Clustering algorithm (variant of k-means for high dimensional dataset)

  • Recommendation System for Yelp dataset
  • Implemented a hybrid recommendation system by combining item based colloborative filtering and model-based colloborative filtering (XGBregressor) to predict ratings of the yelp products. Achieved RMSE of 0.99. Utilized local sensitive hashing technique to optimize the prediction
  • Ankit Mishra
    Fall 2020
    Weenix OS Programming
  • Collaborated with a team of four to program the basic building blocks of an operating system: processes, threads, scheduler and scynchronization primitives (mutex).
  • Designed Virtual File System for managing file descriptor and system calls like open, lookup, read and write.
  • Implemented virtual memory mangement for managing virtual adrress space, page fault handler, fork and copy on write.
  • Ankit Mishra, Pankaj Rajak, Rohan Rout and Akshay Gaur
    Summer 2017

    Publications


  • Full Publication List: Google Scholar

  • 2021
  • Predictive Synthesis of Quantum Materials by Probabilistic Reinforcement Learning

    Pankaj Rajak, Aravind Krishnamoorthy, Ankit Mishra, Rajiv K Kalia, Aiichiro Nakano, Priya Vashishta, Accepted, npj Computational Materials

  • Dielectric Polymer Property Prediction Using Recurrent Neural Networks with Optimizations

    Antonina Nazarova, Liqiu Yang, Kuang Liu, Ankit Mishra, Rajiv Kalia, Ken-ichi Nomura, Aiichiro Nakano, Priya Vashishta, Pankaj Rajak, Accepted, Journal of Chemical Information and Modeling


  • 2020
  • Computational framework for polymer synthesis to study dielectric properties using polarizable reactive molecular dynamics

    Ankit Mishra, Lihua Chen, ZongZe Li, Ken-ichi Nomura, Aravind Krishnamoorthy, Shogo Fukushima, Subodh C Tiwari, Rajiv K Kalia, Aiichiro Nakano, Rampi Ramprasad, Greg Sotzing, Yang Cao, Fuyuki Shimojo, Priya Vashishta, arXiv preprint arXiv:2011.09571
    PDF

  • Atomistic Simulations of Biofouling and Molecular Transfer of Crosslinked Aromatic Polyamide Membrane for Desalination

    Md Symon Jahan Sajib, Ying Wei, Ankit Mishra, Lin Zhang, Ken-ichi Nomura, Rajiv K Kalia, Priya Vashishta, Aiichiro Nakano, Sohail Murad, Tao Wei, Langmuir 2020, 36, 26, 7658–7668
    PDF

  • Evolutionary multi-objective optimization and Pareto-frontal uncertainty quantification of interatomic forcefields for thermal conductivity simulations

    Aravind Krishnamoorthy, Ankit Mishra, Nicholas Grabar, Nitish Baradwaj, Rajiv Kalia, Aiichiro Nakano, Priya Vashishta, Computer Physcis Communications, 2020, 254, 107337
    PDF


  • 2019
  • Hydrogen bond preserving stress release mechanism is key to the resilience of aramid fibers

    Subodh C Tiwari, Kohei Shimamura, Ankit Mishra, Fuyuki Shimojo, Aiichiro Nakano, Rajiv K Kalia, Priya Vashishta, Paulo S Branicio, The Journal of Physical Chemistry B, 2019, 123, 45, 9719-9723
    PDF

  • Materials Genome Software Framework: Scalable Parallel Simulation, Virtual Reality Visualization and Machine Learning

    Ankit Mishra, Nitish Baradwaj, Lindsay Bassman, Brandon K Horton, Subodh Tiwari, Sungwook Hong, Aravind Krishnamoorthy, Erick Moen, Pankaj Rajak, Rajiv Kalia, Aiichiro Nakano, Ken-ichi Nomura, Fuyuki Shimojo, Priya Vashishta, Proceedings of the International Conference on Scientific Computing (CSC), 2019, 125-131
    PDF


  • 2018
  • Free energy of hydration and heat capacity of calcium dipicolinate in Bacillus spore cores

    Ankit Mishra, Aravind Krishnamoorthy, Pankaj Rajak, Subodh Tiwari, Chunyang Sheng, Rajiv K Kalia, Aiichiro Nakano, Priya Vashishta, Applied Physics Letters, 113, 11, 113702
    PDF

  • Multiobjective genetic training and uncertainty quantification of reactive force fields

    Ankit Mishra, Sungwook Hong, Pankaj Rajak, Chunyang Sheng, Ken-ichi Nomura, Rajiv K Kalia, Aiichiro Nakano, Priya Vashishta, npj Computational Materials, 4, 1, 1-7
    PDF

  • Anisotropic frictional heating and defect generation in cyclotrimethylene-trinitramine molecular crystals

    Pankaj Rajak, Ankit Mishra, Chunyang Sheng, Subodh Tiwari, Rajiv K Kalia, Aiichiro Nakano, Priya Vashishta, Applied Physics Letters, 112, 21, 211604
    PDF


  • 2017
  • Gel phase in hydrated calcium dipicolinate

    Pankaj Rajak, Ankit Mishra, Chunyang Sheng, Subodh Tiwari, Aravind Krishnamoorthy, Rajiv K Kalia, Aiichiro Nakano, Priya Vashishta, Applied Physics Letters, 2017, 111, 261, 213701
    PDF


  • 2015
  • Assessment of ground and surface water quality along the river Varuna, Varanasi, India

    Pardeep Singh, RK Chaturvedi, Ankit Mishra, Lata Kumari, Rishikesh Singh, DB Pal, Deen Dayal Giri, Nand Lal Singh, Dhanesh Tiwary, Pradeep Kumar Mishra, Environmental monitoring and assessment, 2017, 187, 4, 170
    PDF


  • Awards


    • Mork Family Department Fellowship, University of Southern California (2014-16)

    • Viterbi Research Assistantship, University of Southern California (2016-2020)

    • Research Article featured on the cover page of nature publishing journal computational materials and Applied Physics Letters

    • Finalist, Maseeh Entrepreneurship Prize Competition, proposed an application (GuidedogAI) for increasing accessibility among visually impaired, University of Southern California 2020

    • Received 99.9 percentile in Joint Entrance Examination (JEE), to receive admission to prestigious Indian Institute of Technology (IIT)