no code implementations • 15 Mar 2024 • Conor M. Artman, Aditya Mate, Ezinne Nwankwo, Aliza Heching, Tsuyoshi Idé, Jiří\, Navrátil, Karthikeyan Shanmugam, Wei Sun, Kush R. Varshney, Lauri Goldkind, Gidi Kroch, Jaclyn Sawyer, Ian Watson
We developed a common algorithmic solution addressing the problem of resource-constrained outreach encountered by social change organizations with different missions and operations: Breaking Ground -- an organization that helps individuals experiencing homelessness in New York transition to permanent housing and Leket -- the national food bank of Israel that rescues food from farms and elsewhere to feed the hungry.
no code implementations • 1 Feb 2024 • Burak Varici, Emre Acartürk, Karthikeyan Shanmugam, Abhishek Kumar, Ali Tajer
The paper addresses both the identifiability and achievability aspects.
1 code implementation • 24 Oct 2023 • Burak Varici, Emre Acartürk, Karthikeyan Shanmugam, Ali Tajer
For identifiability, the paper establishes that perfect recovery of the latent causal model and variables is guaranteed under uncoupled interventions.
no code implementations • 12 Oct 2023 • Shreyas Havaldar, Navodita Sharma, Shubhi Sareen, Karthikeyan Shanmugam, Aravindan Raghuveer
We then use Belief Propagation (BP) to marginalize the Gibbs distribution to obtain pseudo labels.
1 code implementation • 11 Oct 2023 • Shreyas Havaldar, Jatin Chauhan, Karthikeyan Shanmugam, Jay Nandy, Aravindan Raghuveer
Our third contribution is theoretical, where we show that our weighted entropy term along with prediction loss on the training set approximates test loss under covariate shift.
1 code implementation • NeurIPS 2023 • JiaQi Zhang, Chandler Squires, Kristjan Greenewald, Akash Srivastava, Karthikeyan Shanmugam, Caroline Uhler
Causal disentanglement aims to uncover a representation of data using latent variables that are interrelated through a causal model.
no code implementations • 15 Jun 2023 • Shubhada Agrawal, Sandeep Juneja, Karthikeyan Shanmugam, Arun Sai Suggala
Learning paradigms based purely on offline data as well as those based solely on sequential online learning have been well-studied in the literature.
no code implementations • 15 Feb 2023 • Advait Parulekar, Liam Collins, Karthikeyan Shanmugam, Aryan Mokhtari, Sanjay Shakkottai
The goal of contrasting learning is to learn a representation that preserves underlying clusters by keeping samples with similar content, e. g. the ``dogness'' of a dog, close to each other in the space generated by the representation.
no code implementations • 19 Jan 2023 • Burak Varici, Emre Acarturk, Karthikeyan Shanmugam, Abhishek Kumar, Ali Tajer
The objectives are: (i) recovering the unknown linear transformation (up to scaling) and (ii) determining the directed acyclic graph (DAG) underlying the latent variables.
no code implementations • 17 Jan 2023 • Soumyabrata Pal, Arun Sai Suggala, Karthikeyan Shanmugam, Prateek Jain
Instead, we propose LATTICE (Latent bAndiTs via maTrIx ComplEtion) which allows exploitation of the latent cluster structure to provide the minimax optimal regret of $\widetilde{O}(\sqrt{(\mathsf{M}+\mathsf{N})\mathsf{T}})$, when the number of clusters is $\widetilde{O}(1)$.
no code implementations • 12 Dec 2022 • Nishant Jain, Karthikeyan Shanmugam, Pradeep Shenoy
Predictive uncertainty-a model's self awareness regarding its accuracy on an input-is key for both building robust models via training interventions and for test-time applications such as selective classification.
1 code implementation • 26 Aug 2022 • Burak Varici, Karthikeyan Shanmugam, Prasanna Sattigeri, Ali Tajer
Two linear mechanisms, one soft intervention and one observational, are assumed for each node, giving rise to $2^N$ possible interventions.
no code implementations • 14 Jul 2022 • Samuel C. Hoffman, Kahini Wadhawan, Payel Das, Prasanna Sattigeri, Karthikeyan Shanmugam
In this work, we provide a simple algorithm that relies on perturbation experiments on latent codes of a pre-trained generative autoencoder to uncover a causal graph that is implied by the generative model.
no code implementations • 30 May 2022 • Advait Parulekar, Karthikeyan Shanmugam, Sanjay Shakkottai
These are representations of the covariates such that the best model on top of the representation is invariant across training environments.
no code implementations • 8 Feb 2022 • Hamid Dadkhahi, Jesus Rios, Karthikeyan Shanmugam, Payel Das
In order to improve the performance and sample efficiency of such algorithms, we propose to use existing methods in conjunction with a surrogate model for the black-box evaluations over purely categorical variables.
1 code implementation • 2 Feb 2022 • Keerthiram Murugesan, Vijay Sadashivaiah, Ronny Luss, Karthikeyan Shanmugam, Pin-Yu Chen, Amit Dhurandhar
Knowledge transfer between heterogeneous source and target networks and tasks has received a lot of attention in recent times as large amounts of quality labeled data can be difficult to obtain in many applications.
no code implementations • NeurIPS 2021 • Isha Puri, Amit Dhurandhar, Tejaswini Pedapati, Karthikeyan Shanmugam, Dennis Wei, Kush R. Varshney
We experiment on nonlinear synthetic functions and are able to accurately model as well as estimate feature attributions and even higher order terms in some cases, which is a testament to the representational power as well as interpretability of such architectures.
1 code implementation • NeurIPS 2021 • Burak Varici, Karthikeyan Shanmugam, Prasanna Sattigeri, Ali Tajer
This paper considers the problem of estimating the unknown intervention targets in a causal directed acyclic graph from observational and interventional data.
no code implementations • ICLR 2022 • Keerthiram Murugesan, Vijay Sadashivaiah, Ronny Luss, Karthikeyan Shanmugam, Pin-Yu Chen, Amit Dhurandhar
Knowledge transfer between heterogeneous source and target networks and tasks has received a lot of attention in recent times as large amounts of quality labelled data can be difficult to obtain in many applications.
no code implementations • 24 Sep 2021 • Vijay Arya, Rachel K. E. Bellamy, Pin-Yu Chen, Amit Dhurandhar, Michael Hind, Samuel C. Hoffman, Stephanie Houde, Q. Vera Liao, Ronny Luss, Aleksandra Mojsilovic, Sami Mourad, Pablo Pedemonte, Ramya Raghavendra, John Richards, Prasanna Sattigeri, Karthikeyan Shanmugam, Moninder Singh, Kush R. Varshney, Dennis Wei, Yunfeng Zhang
As artificial intelligence and machine learning algorithms become increasingly prevalent in society, multiple stakeholders are calling for these algorithms to provide explanations.
no code implementations • 7 Jul 2021 • Nihal Sharma, Soumya Basu, Karthikeyan Shanmugam, Sanjay Shakkottai
The agent interacts with the environment over episodes, with each episode having different context distributions; this results in the `best expert' changing across episodes.
no code implementations • NeurIPS 2021 • Zaiwei Chen, Siva Theja Maguluri, Sanjay Shakkottai, Karthikeyan Shanmugam
Our key step is to show that the generalized Bellman operator is simultaneously a contraction mapping with respect to a weighted $\ell_p$-norm for each $p$ in $[1,\infty)$, with a common contraction factor.
1 code implementation • 22 Jun 2021 • Abhin Shah, Karthikeyan Shanmugam, Kartik Ahuja
Our main result strengthens these prior results by showing that under a different expert-driven structural knowledge -- that one variable is a direct causal parent of treatment variable -- remarkably, testing for subsets (not involving the known parent variable) that are valid back-doors is equivalent to an invariance test.
2 code implementations • 13 Mar 2021 • Abhin Shah, Kartik Ahuja, Karthikeyan Shanmugam, Dennis Wei, Kush Varshney, Amit Dhurandhar
Inferring causal individual treatment effect (ITE) from observational data is a challenging problem whose difficulty is exacerbated by the presence of treatment assignment bias.
no code implementations • 5 Mar 2021 • Kanthi Sarpatwar, Karthik Nandakumar, Nalini Ratha, James Rayfield, Karthikeyan Shanmugam, Sharath Pankanti, Roman Vaculin
In this work, we propose a framework to transfer knowledge extracted by complex decision tree ensembles to shallow neural networks (referred to as DTNets) that are highly conducive to encrypted inference.
no code implementations • 2 Feb 2021 • Zaiwei Chen, Siva Theja Maguluri, Sanjay Shakkottai, Karthikeyan Shanmugam
As a by-product, by analyzing the convergence bounds of $n$-step TD and TD$(\lambda)$, we provide theoretical insights into the bias-variance trade-off, i. e., efficiency of bootstrapping in RL.
1 code implementation • NeurIPS 2020 • Chandler Squires, Sara Magliacane, Kristjan Greenewald, Dmitriy Katz, Murat Kocaoglu, Karthikeyan Shanmugam
Most existing works focus on \textit{worst-case} or \textit{average-case} lower bounds for the number of interventions required to orient a DAG.
no code implementations • NeurIPS 2020 • Amin Jaber, Murat Kocaoglu, Karthikeyan Shanmugam, Elias Bareinboim
One fundamental problem in the empirical sciences is of reconstructing the causal structure that underlies a phenomenon of interest through observation and experimentation.
no code implementations • 2 Nov 2020 • Advait Parulekar, Soumya Basu, Aditya Gopalan, Karthikeyan Shanmugam, Sanjay Shakkottai
We study a variant of the stochastic linear bandit problem wherein we optimize a linear objective function but rewards are accrued only orthogonal to an unknown subspace (which we interpret as a \textit{protected space}) given only zero-order stochastic oracle access to both the objective itself and protected subspace.
4 code implementations • 1 Nov 2020 • Chandler Squires, Sara Magliacane, Kristjan Greenewald, Dmitriy Katz, Murat Kocaoglu, Karthikeyan Shanmugam
Most existing works focus on worst-case or average-case lower bounds for the number of interventions required to orient a DAG.
3 code implementations • ICLR 2021 • Kartik Ahuja, Jun Wang, Amit Dhurandhar, Karthikeyan Shanmugam, Kush R. Varshney
Recently, invariant risk minimization (IRM) was proposed as a promising solution to address out-of-distribution (OOD) generalization.
3 code implementations • 28 Oct 2020 • Kartik Ahuja, Karthikeyan Shanmugam, Amit Dhurandhar
In Ahuja et al., it was shown that solving for the Nash equilibria of a new class of "ensemble-games" is equivalent to solving IRM.
no code implementations • 10 Jun 2020 • Sainyam Galhotra, Karthikeyan Shanmugam, Prasanna Sattigeri, Kush R. Varshney
In this work, we consider fairness in the integration component of data management, aiming to identify features that improve prediction without adding any bias to the dataset.
no code implementations • 6 Jun 2020 • Hamid Dadkhahi, Karthikeyan Shanmugam, Jesus Rios, Payel Das, Samuel Hoffman, Troy David Loeffler, Subramanian Sankaranarayanan
We consider the problem of black-box function optimization over the boolean hypercube.
no code implementations • 21 Feb 2020 • Tian Gao, Dharmashankar Subramanian, Karthikeyan Shanmugam, Debarun Bhattacharjya, Nicholas Mattei
Event datasets are sequences of events of various types occurring irregularly over the time-line, and they are increasingly prevalent in numerous domains.
no code implementations • 19 Feb 2020 • Nihal Sharma, Soumya Basu, Karthikeyan Shanmugam, Sanjay Shakkottai
We study a variant of the multi-armed bandit problem where side information in the form of bounds on the mean of each arm is provided.
no code implementations • NeurIPS 2020 • Tejaswini Pedapati, Avinash Balakrishnan, Karthikeyan Shanmugam, Amit Dhurandhar
Based on a key insight we propose a novel method where we create custom boolean features from sparse local contrastive explanations of the black-box model and then train a globally transparent model on just these, and showcase empirically that such models have higher local consistency compared with other known strategies, while still being close in performance to models that are trained with access to the original data.
3 code implementations • ICML 2020 • Kartik Ahuja, Karthikeyan Shanmugam, Kush R. Varshney, Amit Dhurandhar
The standard risk minimization paradigm of machine learning is brittle when operating in environments whose test distributions are different from the training distribution due to spurious correlations.
no code implementations • NeurIPS 2020 • Zaiwei Chen, Siva Theja Maguluri, Sanjay Shakkottai, Karthikeyan Shanmugam
In particular, we use it to establish the first-known convergence rate of the V-trace algorithm for off-policy TD-learning.
no code implementations • NeurIPS 2019 • Kristjan Greenewald, Dmitriy Katz, Karthikeyan Shanmugam, Sara Magliacane, Murat Kocaoglu, Enric Boix Adsera, Guy Bresler
We consider the problem of experimental design for learning causal graphs that have a tree structure.
no code implementations • NeurIPS 2019 • Murat Kocaoglu, Amin Jaber, Karthikeyan Shanmugam, Elias Bareinboim
We introduce a novel notion of interventional equivalence class of causal graphs with latent variables based on these invariances, which associates each graphical structure with a set of interventional distributions that respect the do-calculus rules.
no code implementations • NeurIPS 2019 • Kanthi Sarpatwar, Karthikeyan Shanmugam, Venkata Sitaramagiridharganesh Ganapavarapu, Ashish Jagmohan, Roman Vaculin
Our central result is a novel protocol that (a) ensures the curator accesses at most $O(K^{\frac{1}{3}}|D_s| + |D_v|)$ points (b) has formal privacy guarantees on the leakage of information between the data owners and (c) closely matches the best known non-private greedy algorithm.
no code implementations • 25 Sep 2019 • Amit Dhurandhar, Karthikeyan Shanmugam, Ronny Luss
Our method also leverages the per sample hardness estimate of the simple model which is not the case with the prior works which primarily consider the complex model's confidences/predictions and is thus conceptually novel.
2 code implementations • 6 Sep 2019 • Vijay Arya, Rachel K. E. Bellamy, Pin-Yu Chen, Amit Dhurandhar, Michael Hind, Samuel C. Hoffman, Stephanie Houde, Q. Vera Liao, Ronny Luss, Aleksandra Mojsilović, Sami Mourad, Pablo Pedemonte, Ramya Raghavendra, John Richards, Prasanna Sattigeri, Karthikeyan Shanmugam, Moninder Singh, Kush R. Varshney, Dennis Wei, Yunfeng Zhang
Equally important, we provide a taxonomy to help entities requiring explanations to navigate the space of explanation methods, not only those in the toolkit but also in the broader literature on explainability.
1 code implementation • NeurIPS 2020 • Matthew Faw, Rajat Sen, Karthikeyan Shanmugam, Constantine Caramanis, Sanjay Shakkottai
We consider a covariate shift problem where one has access to several different training datasets for the same learning problem and a small validation set which possibly differs from all the individual training distributions.
no code implementations • 31 May 2019 • Amit Dhurandhar, Tejaswini Pedapati, Avinash Balakrishnan, Pin-Yu Chen, Karthikeyan Shanmugam, Ruchir Puri
Recently, a method [7] was proposed to generate contrastive explanations for differentiable models such as deep neural networks, where one has complete access to the model.
no code implementations • ICML 2020 • Amit Dhurandhar, Karthikeyan Shanmugam, Ronny Luss
Our method also leverages the per sample hardness estimate of the simple model which is not the case with the prior works which primarily consider the complex model's confidences/predictions and is thus conceptually novel.
2 code implementations • 29 May 2019 • Ronny Luss, Pin-Yu Chen, Amit Dhurandhar, Prasanna Sattigeri, Yunfeng Zhang, Karthikeyan Shanmugam, Chun-Chen Tu
As the application of deep neural networks proliferates in numerous areas such as medical imaging, video surveillance, and self driving cars, the need for explaining the decisions of these models has become a hot research topic, both at the global and local level.
no code implementations • 5 Mar 2019 • Dmitriy Katz, Karthikeyan Shanmugam, Chandler Squires, Caroline Uhler
For constant density, we show that the expected $\log$ observational MEC size asymptotically (in the number of vertices) approaches a constant.
no code implementations • NeurIPS 2018 • Amit Dhurandhar, Karthikeyan Shanmugam, Ronny Luss, Peder Olsen
Our transfer method involves a theoretically justified weighting of samples during the training of the simple model using confidence scores of these intermediate layers.
1 code implementation • 25 Jun 2018 • Rajat Sen, Karthikeyan Shanmugam, Himanshu Asnani, Arman Rahimzamani, Sreeram Kannan
Given independent samples generated from the joint distribution $p(\mathbf{x},\mathbf{y},\mathbf{z})$, we study the problem of Conditional Independence (CI-Testing), i. e., whether the joint equals the CI distribution $p^{CI}(\mathbf{x},\mathbf{y},\mathbf{z})= p(\mathbf{z}) p(\mathbf{y}|\mathbf{z})p(\mathbf{x}|\mathbf{z})$ or not.
no code implementations • 24 May 2018 • Bernat Guillen Pegueroles, Bhanukiran Vinzamuri, Karthikeyan Shanmugam, Steve Hedden, Jonathan D. Moyer, Kush R. Varshney
Almost all existing Granger causal algorithms condition on a large number of variables (all but two variables) to test for effects between a pair of variables.
no code implementations • 14 May 2018 • Tongfei Chen, Jiří Navrátil, Vijay Iyengar, Karthikeyan Shanmugam
We propose a novel confidence scoring mechanism for deep neural networks based on a two-model paradigm involving a base model and a meta-model.
1 code implementation • 23 Feb 2018 • Rajat Sen, Karthikeyan Shanmugam, Nihal Sharma, Sanjay Shakkottai
We consider the problem of contextual bandits with stochastic experts, which is a variation of the traditional stochastic contextual bandit with experts problem.
4 code implementations • NeurIPS 2018 • Amit Dhurandhar, Pin-Yu Chen, Ronny Luss, Chun-Chen Tu, Pai-Shun Ting, Karthikeyan Shanmugam, Payel Das
important object pixels in an image) to justify its classification and analogously what should be minimally and necessarily \emph{absent} (viz.
no code implementations • NeurIPS 2017 • Murat Kocaoglu, Karthikeyan Shanmugam, Elias Bareinboim
Next, we propose an algorithm that uses only O(d^2 log n) interventions that can learn the latents between both non-adjacent and adjacent variables.
1 code implementation • NeurIPS 2017 • Rajat Sen, Ananda Theertha Suresh, Karthikeyan Shanmugam, Alexandros G. Dimakis, Sanjay Shakkottai
We consider the problem of non-parametric Conditional Independence testing (CI testing) for continuous random variables.
no code implementations • 12 Jul 2017 • Amit Dhurandhar, Vijay Iyengar, Ronny Luss, Karthikeyan Shanmugam
We provide a novel notion of what it means to be interpretable, looking past the usual association with human understanding.
no code implementations • 9 Jun 2017 • Amit Dhurandhar, Vijay Iyengar, Ronny Luss, Karthikeyan Shanmugam
This leads to the insight that the improvement in the target model is not only a function of the oracle model's performance, but also its relative complexity with respect to the target model.
no code implementations • 8 Mar 2017 • Karthikeyan Shanmugam, Murat Kocaoglu, Alexandros G. Dimakis, Sujay Sanghavi
We consider support recovery in the quadratic logistic regression setting - where the target depends on both p linear terms $x_i$ and up to $p^2$ quadratic terms $x_i x_j$.
no code implementations • ICML 2017 • Rajat Sen, Karthikeyan Shanmugam, Alexandros G. Dimakis, Sanjay Shakkottai
Motivated by applications in computational advertising and systems biology, we consider the problem of identifying the best out of several possible soft interventions at a source node $V$ in an acyclic causal directed graph, to maximize the expected value of a target node $Y$ (located downstream of $V$).
no code implementations • 1 Jun 2016 • Rajat Sen, Karthikeyan Shanmugam, Murat Kocaoglu, Alexandros G. Dimakis, Sanjay Shakkottai
Our algorithm achieves a regret of $\mathcal{O}\left(L\mathrm{poly}(m, \log K) \log T \right)$ at time $T$, as compared to $\mathcal{O}(LK\log T)$ for conventional contextual bandits, assuming a constant gap between the best arm and the rest for each context.
2 code implementations • NeurIPS 2015 • Karthikeyan Shanmugam, Murat Kocaoglu, Alexandros G. Dimakis, Sriram Vishwanath
We prove that any deterministic adaptive algorithm needs to be a separating system in order to learn complete graphs in the worst case.
no code implementations • NeurIPS 2014 • Karthikeyan Shanmugam, Rashish Tandon, Alexandros G. Dimakis, Pradeep Ravikumar
We provide a general framework for computing lower-bounds on the sample complexity of recovering the underlying graphs of Ising models, given i. i. d samples.
no code implementations • NeurIPS 2014 • Murat Kocaoglu, Karthikeyan Shanmugam, Alexandros G. Dimakis, Adam Klivans
We give an algorithm for exactly reconstructing f given random examples from the uniform distribution on $\{-1, 1\}^n$ that runs in time polynomial in $n$ and $2s$ and succeeds if the function satisfies the unique sign property: there is one output value which corresponds to a unique set of values of the participating parities.