Shashank Singh

Jawaharlal Institute of Post Graduate Medical Education and Research

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Barnabás Póczos

Carnegie Mellon University

Amrita Chaturvedi

Indian Institute of Technology BHU

Ananya Uppal

PES University

Jaeah Kim

Carnegie Mellon University

Erik D. Thiessen

Carnegie Mellon University

Anna V. Fisher

Carnegie Mellon University

Chun‐Liang Li

St. Jude Children's Research Hospital

Tomas Pfister

University of Oxford

Vikas Menon

Jawaharlal Institute of Post Graduate Medical Education and Research

Arkady Epshteyn

Google (United States)

Cooperative Institutions

Google (United States)

Carnegie Mellon University

Post Graduate Institute of Medical Education and Research

Banaras Hindu University

University of Pittsburgh

Shibuya (Japan)

Indian Institute of Technology BHU

Sher-i-Kashmir Institute of Medical Sciences

Institute of Medical Sciences

Maulana Azad Medical College

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Finite-Sample Analysis of Fixed-k Nearest Neighbor Density Functional Estimators

Neural Information Processing Systems (2016)

Shashank Singh Barnabás Póczos

We provide finite-sample analysis of a general framework for using k-nearest neighbor statistics to estimate functionals of a nonparametric continuous probability density, including entropies and divergences. Rather than plugging a consistent density estimate (which requires k → ∞ as the sample size n → ∞) into the functional of interest, the estimators we consider fix k and perform a bias correction. This can be more efficient computationally, and, as we show, statistically, leading to faster convergence rates. Our framework unifies several previous estimators, for most of which ours are the first finite sample guarantees.

Sample (material)

Density estimation

Source

Cite

Citations (29)

Nonparanormal Information Estimation

arXiv (Cornell University) (2017)

Shashank Singh Barnabás Póczos

We study the problem of using i.i.d. samples from an unknown multivariate probability distribution $p$ to estimate the mutual information of $p$. This problem has recently received attention in two settings: (1) where $p$ is assumed to be Gaussian and (2) where $p$ is assumed only to lie in a large nonparametric smoothness class. Estimators proposed for the Gaussian case converge in high dimensions when the Gaussian assumption holds, but are brittle, failing dramatically when $p$ is not Gaussian. Estimators proposed for the nonparametric case fail to converge with realistic sample sizes except in very low dimensions. As a result, there is a lack of robust mutual information estimators for many realistic data. To address this, we propose estimators for mutual information when $p$ is assumed to be a nonparanormal (a.k.a., Gaussian copula) model, a semiparametric compromise between Gaussian and nonparametric extremes. Using theoretical bounds and experiments, we show these estimators strike a practical balance between robustness and scaling with dimensionality.

Smoothness

Source

Cite

Citations (7)

Handling Real-Word Errors of Hindi Language using N-gram and Confusion Set

Shashank Singh Shailendra Singh

The two major typographic errors of any language are non-word errors and real-word errors. The researchers have worked rigorously for the former error but the latter has not been given the much attention all this while. In this paper, a technique to identify the real-word error of Hindi language has been proposed which combines the bigram, trigram and confusion set (CS) methods. Left bigrams, right bigrams and trigrams are calculated by taking in to account the immediate left, immediate right word with the Hindi test word and the immediate left, the Hindi test word and the immediate right word respectively. A group of most confusable words is created using Levenstein edit distance method. After that, a composite score is calculated for all the members of the CS using bigram and trigram probabilities. The calculated composite score is used to prepare the suggestion list for the erroneous word. A Hindi text file of 2000 words has been used to evaluate and verify the proposed method which offers considerably good results. It gives the precision, recall and F-score as .70-.75, .80-.85 and .70-.80 respectively. In future, the research can be done to improve the results by considering the whole sentence at once.

Bigram

n-gram

10.1109/aicai.2019.8701394

Cite

Citations (4)

Robust Density Estimation under Besov IPM Losses

Neural Information Processing Systems (2020)

Ananya Uppal Shashank Singh Barnabás Póczos

We study minimax convergence rates of nonparametric density estimation in the Huber contamination model, in which a proportion of the data comes from an unknown outlier distribution. We provide the first results for this problem under a large family of losses, called Besov integral probability metrics (IPMs), that includes $\mathcal{L}^p$, Wasserstein, Kolmogorov-Smirnov, and other common distances between probability distributions. Specifically, under a range of smoothness assumptions on the population and outlier distributions, we show that a re-scaled thresholding wavelet series estimator achieves minimax optimal convergence rates under a wide variety of losses. Finally, based on connections that have recently been shown between nonparametric density estimation under IPM losses and generative adversarial networks (GANs), we show that certain GAN architectures also achieve these minimax rates.

Smoothness

Density estimation

Source

Cite

Citations (0)

Measuring Selective Sustained Attention in Children with TrackIt and Eyetracking.

Cognitive Science (2019)

Jaeah Kim Shashank Singh Emily Keebler Erik D. Thiessen Anna V. Fisher

Source

Cite

Citations (0)

A prospective evaluation of AI-augmented epidemiology to forecast COVID-19 in the USA and Japan

npj Digital Medicine (2021)

Sercan Ö. Arık Joel Shor Rajarishi Sinha Jinsung Yoon Joseph R. Ledsam

Abstract The COVID-19 pandemic has highlighted the global need for reliable models of disease spread. We propose an AI-augmented forecast modeling framework that provides daily predictions of the expected number of confirmed COVID-19 deaths, cases, and hospitalizations during the following 4 weeks. We present an international, prospective evaluation of our models’ performance across all states and counties in the USA and prefectures in Japan. Nationally, incident mean absolute percentage error (MAPE) for predicting COVID-19 associated deaths during prospective deployment remained consistently <8% (US) and <29% (Japan), while cumulative MAPE remained <2% (US) and <10% (Japan). We show that our models perform well even during periods of considerable change in population behavior, and are robust to demographic differences across different geographic locations. We further demonstrate that our framework provides meaningful explanatory insights with the models accurately adapting to local and national policy interventions. Our framework enables counterfactual simulations, which indicate continuing Non-Pharmaceutical Interventions alongside vaccinations is essential for faster recovery from the pandemic, delaying the application of interventions has a detrimental effect, and allow exploration of the consequences of different vaccination strategies. The COVID-19 pandemic remains a global emergency. In the face of substantial challenges ahead, the approach presented here has the potential to inform critical decisions.

Pandemic

10.1038/s41746-021-00511-7

Cite

Citations (25)

Nonparametric Density Estimation under Adversarial Losses

Neural Information Processing Systems (2018)

Shashank Singh Ananya Uppal Boyue Li Chun‐Liang Li Manzil Zaheer

We study minimax convergence rates of nonparametric density estimation under a large class of loss functions called ``adversarial losses'', which, besides classical L^p losses, includes maximum mean discrepancy (MMD), Wasserstein distance, and total variation distance. These losses are closely related to the losses encoded by discriminator networks in generative adversarial networks (GANs). In a general framework, we study how the choice of loss and the assumed smoothness of the underlying density together determine the minimax rate. We also discuss implications for training GANs based on deep ReLU networks, and more general connections to learning implicit generative models in a minimax statistical sense.

Smoothness

Discriminator

Density estimation

Variation (astronomy)

Source

Cite

Citations (18)

Applying Extreme Gradient Boosting for Surface EMG Based Sign Language Recognition

Lecture notes in networks and systems (2021)

Shashank Singh Amrita Chaturvedi Alok Prakash

Boosting

Gradient boosting

American Sign Language

10.1007/978-3-030-82469-3_16

Cite

Citations (2)

A bacterial based distributed gradient descent model for mass scale evacuations

Swarm and Evolutionary Computation (2019)

Sabrina Mohd Rashid Shashank Singh Saket Navlakha Ziv Bar−Joseph

10.1016/j.swevo.2019.01.011

Cite

Citations (4)

Nonparametric Density Estimation under Besov IPM Losses.

arXiv (Cornell University) (2019)

Ananya Uppal Shashank Singh Barnabás Póczos

We study the problem of estimating a nonparametric probability density under a large family of losses called Besov IPMs, which include, for example, $\mathcal{L}^p$ distances, total variation distance, and generalizations of both Wasserstein and Kolmogorov-Smirnov distances. For a wide variety of settings, we provide both lower and upper bounds, identifying precisely how the choice of loss function and assumptions on the data interact to determine the minimax optimal convergence rate. We also show that linear distribution estimates, such as the empirical distribution or kernel density estimator, often fail to converge at the optimal rate. Our bounds generalize, unify, or improve several recent and classical results. Moreover, IPMs can be used to formalize a statistical model of generative adversarial networks (GANs). Thus, we show how our results imply bounds on the statistical error of a GAN, showing, for example, that GANs can strictly outperform the best linear estimator.

Kernel density estimation

Kernel (algebra)

Empirical distribution function

Density estimation

Source

Cite

Citations (5)