Home > People > Amal Agarwal
Amal Agarwal

Amal Agarwal

Main Content

418 Thomas Building
University Park, PA 16802
Phone: (814) 441-2119


  1. Dual Degree (B.Tech.+ M.Tech.) in Engineering Physics with Minor in Statistics
  2. Indian Institute of Technology (IIT) Bombay
  3. August 2014


If you are here, thanks for checking out my profile! Currently, I am a 5th-year graduate research assistant at the Department of Statistics at Pennsylvania State University. Before coming to Penn State, I graduated with a Masters in Engineering Physics from IIT Bombay.

I am Newtonian as a physicist and Bayesian as a statistician. I see Newton’s third law of action-reaction everywhere. As a direct reaction to this observation, I like to experiment with different ideas. Some work and many don’t. I fail fast and I learn. All these failures become priors in future trials which allows me to update the posterior of my life model. And what we do in this life echoes in eternity, isn’t it? (and yes, that is from Gladiator :-))

Throughout my short and relatively meaningless existence, I have learned enough to know I know nothing. Well, a.s. (almost surely), if you know what I mean :-)). I consider myself a gritty problem solver. I like to think backward from applications and therefore I often find myself playing with data at the onset of a project. All of my research endeavors are motivated by challenging real-life data problems.

I consider meta-learning, i.e. learning how to learn, to be the most important life philosophy. With a good handle on this, I believe ideas can grow exponentially as nodes on a big knowledge network.    

To know more about me, feel free to check out my personal website. Don’t hesitate to shoot me an email at amalag.19@gmail.com, if you'd like to explore/discuss ideas together and/or your research interests match with mine and you wish to collaborate. Just one match can create an explosion! :-))

Selected Publications

In this section, I attempt to describe my research portfolio by segregating papers hierarchically based on topics and applications within each topic. At the beginning of each topic, I list rigorously peer-reviewed/under revision/under review, full-length papers at reputable journals or conferences. Towards the end, I mention currently working papers. Please note that I exclude all short workshop/conference papers, poster/talk abstracts, and papers with (foreseeable) little impact/no review.

  • Large Scale Network Clustering Papers (Applied ML):
    • Applications to Detecting Polluters in River Networks:
    • Applications to Recommender Systems in Marketing:
      • Guided by Your Stars: Two-Mode Segmentation Using Large-Scale Online Product Rating Networks 
        Amal Agarwal (co-first author)Qian Chen (co-first author), Duncan Fong, and Wayne DeSarbo.
        Under review in Marketing Science.
    • Applications to Dynamically evolving Networks:
      • Temporal Exponential-Family Random Graph Models with Time-Evolving Latent Block Structure for Dynamic Networks
        Amal Agarwal
        , Kevin Lee and Lingzhou Xue.
        Under review in Journal of Business and Economic Statistics.
      • Semiparametric Mixture of Exponential-Family Random Graph Models with Dynamically Varying Parameters (working paper)
        Kevin Lee, Amal Agarwal and Lingzhou Xue.
  • Geoscience Papers:
    • Applications of nonparametric solutions of Behrens-Fisher problem:
    • Applications to automated testing in big river networks (working paper):
      • GeoNet: An automated geochemical network analysis with application to detecting stream water contamination
  • Causal Inference:
    • Applications of transfer function models to time series data:
      • Discovery of Causal Time Intervals
        Zhenhui Li, Guanjie Zheng, Amal Agarwal, and Lingzhou Xue. 
        SDM’17: the Seventeenth SIAM International Conference on Data Mining, 804-812, 2017.
  • Bayesian Particle Filter Papers:
    • Applications to Acoustics and Signal Processing:

Research Interests

Currently, I am working on several exciting research topics under the advisorship of Dr. Lingzhou Xue.

Primary Research Interests:

  1. Large-scale network analysis
    • Time-Evolving Community Detection in Dynamic Networks
    • Non-Parametric Clustering of Continuous Weighted Networks
    • Bipartite clustering with an embedded proportional odds model for discrete weights
  2. Deep Learning through Knowledge Enhanced Neural Networks (DL-KENN)
  3. Statistical Machine Learning
  4. Variational Inference
  5. Stochastic Optimization
  6. Bayesian Analysis and MCMC algorithms
  7. Parallel Computing and Visualization

Collaborative Research Interests:

  1. Geosciences-
    • Network Applications in sparse spatiotemporal Environmental Big Data
    • Building Polluter Detection tools using R Shiny and Leaflet (see GeoNet Application)
  2. IBM Research- 
    • Applications of DL-KENN to healthcare data (ASCVD risk in diabetes patients).
  3. Business-
    • Recommender Systems via Two-Mode Segmentation of Consumer-Product Review Networks
  4. Genomics-
    • Constrained Penalized Regression Models for big chromosome matrices (HiC Data)

Honors and Awards

  • Insight Data Science Fellow.
  • Student Travel Scholarship to 36th ASA QPRC.
  • J. Keith Ord Scholarship in Statistics.
  • Best Student Paper Award in Risk Analysis, ASA.
  • Student Travel Award and Finalist in student paper competition in IISA.

Navigation for this Section