arXiv subjects

Lav R. Varshney

Publications and source records attributed to Lav R. Varshney.

At least 91 records · Page 5Linked to original sources

arXiv2022

Accelerated Design and Deployment of Low-Carbon Concrete for Data Centers

Concrete is the most widely used engineered material in the world with more than 10 billion tons produced annually. Unfortunately, with that scale comes a significant burden in terms of energy, water, and release of greenhouse gases and other pollutants; indeed 8% of worldwide carbon emissions are attributed to the production of cement, a key ingredient in concrete. As such, there is interest in creating concrete formulas that minimize this environmental burden, while satisfying engineering performance requirements including compressive strength. Specifically for computing, concrete is a major ingredient in the construction of data centers. In this work, we use conditional variational autoencoders (CVAEs), a type of semi-supervised generative artificial intelligence (AI) model, to discover concrete formulas with desired properties. Our model is trained just using a small open dataset from the UCI Machine Learning Repository joined with environmental impact data from standard lifecycle analysis. Computational predictions demonstrate CVAEs can design concrete formulas with much lower carbon requirements than existing formulations while meeting design requirements. Next we report laboratory-based compressive strength experiments for five AI-generated formulations, which demonstrate that the formulations exceed design requirements. The resulting formulations were then used by Ozinga Ready Mix -- a concrete supplier -- to generate field-ready concrete formulations, based on local conditions and their expertise in concrete design. Finally, we report on how these formulations were used in the construction of buildings and structures in a Meta data center in DeKalb, IL, USA. Results from field experiments as part of this real-world deployment corroborate the efficacy of AI-generated low-carbon concrete mixes.

cs.AI↗

arXiv2022

The CEO Problem with $r$th Power of Difference and Logarithmic Distortions

The CEO problem has received much attention since first introduced by Berger et al., but there are limited results on non-Gaussian models with non-quadratic distortion measures. In this work, we extend the quadratic Gaussian CEO problem to two non-Gaussian settings with general $r$th power of difference distortion. Assuming an identical observation channel across agents, we study the asymptotics of distortion decay as the number of agents and sum-rate, $R_{sum}$, grow without bound, while individual rates vanish. The first setting is a regular source-observation model with $r$th power of difference distortion, which subsumes the quadratic Gaussian CEO problem, and we establish that the distortion decays at $\mathcal{O}(R_{sum}^{-r/2})$ when $r \ge 2$. We use sample median estimation after the Berger-Tung scheme for achievability. The other setting is a \emph{non-regular} source-observation model, including uniform additive noise models, with $r$th power of difference distortion for which estimation-theoretic regularity conditions do not hold. The distortion decay $\mathcal{O}(R_{sum}^{-r})$ when $r \ge 1$ is obtained for the non-regular model by midrange estimator following the Berger-Tung scheme. We also provide converses based on the Shannon lower bound for the regular model and the Chazan-Zakai-Ziv bound for the non-regular model, respectively. Lastly, we provide a sufficient condition for the regular model, under which quadratic and logarithmic distortions are asymptotically equivalent by an entropy power relationship as the number of agents grows. This proof relies on the Bernstein-von Mises theorem.

cs.IT↗

arXiv2022

Advanced Methods for Connectome-Based Predictive Modeling of Human Intelligence: A Novel Approach Based on Individual Differences in Cortical Topography

Individual differences in human intelligence can be modeled and predicted from in vivo neurobiological connectivity. Many established modeling frameworks for predicting intelligence, however, discard higher-order information about individual differences in brain network topology, and show only moderate performance when generalized to make predictions in out-of-sample subjects. In this paper, we propose that connectome-based predictive modeling, a common predictive modeling framework for neuroscience data, can be productively modified to incorporate information about brain network topology and individual differences via the incorporation of bagged decision trees and the network based statistic. These modifications produce a novel predictive modeling framework that leverages individual differences in cortical tractography to generate accurate regression predictions of intelligence scores. Network topology-based feature selection provides for natively interpretable networks as input features, increasing the model's explainability. Investigating the proposed modeling framework's efficacy, we find that advanced connectome-based predictive modeling generates neuroscience predictions that account for a significantly greater proportion of variance in general intelligence scores than previously established methods, advancing our scientific understanding of the network architecture that underlies human intelligence.

q-bio.NC↗

arXiv2021

Balancing Fairness and Robustness via Partial Invariance

The Invariant Risk Minimization (IRM) framework aims to learn invariant features from a set of environments for solving the out-of-distribution (OOD) generalization problem. The underlying assumption is that the causal components of the data generating distributions remain constant across the environments or alternately, the data "overlaps" across environments to find meaningful invariant features. Consequently, when the "overlap" assumption does not hold, the set of truly invariant features may not be sufficient for optimal prediction performance. Such cases arise naturally in networked settings and hierarchical data-generating models, wherein the IRM performance becomes suboptimal. To mitigate this failure case, we argue for a partial invariance framework. The key idea is to introduce flexibility into the IRM framework by partitioning the environments based on hierarchical differences, while enforcing invariance locally within the partitions. We motivate this framework in classification settings with causal distribution shifts across environments. Our results show the capability of the partial invariant risk minimization to alleviate the trade-off between fairness and risk in certain settings.

cs.LG↗

arXiv2021

Expected Extinction Times of Epidemics with State-Dependent Infectiousness

We model an epidemic where the per-person infectiousness in a network of geographic localities changes with the total number of active cases. This would happen as people adopt more stringent non-pharmaceutical precautions when the population has a larger number of active cases. We show that there exists a sharp threshold such that when the curing rate for the infection is above this threshold, the mean time for the epidemic to die out is logarithmic in the initial infection size, whereas when the curing rate is below this threshold, the mean time for epidemic extinction is infinite. We also show that when the per-person infectiousness goes to zero asymptotically as a function of the number of active cases, the mean extinction times all have the same asymptote independent of network structure. Simulations bear out these results, while also demonstrating that if the per-person infectiousness is large when the epidemic size is small (i.e., the precautions are lax when the epidemic is small and only get stringent after the epidemic has become large), it might take a very long time for the epidemic to die out. We also provide some analytical insight into these observations.

cs.SI↗

arXiv2021

Decision Making in Star Networks with Incorrect Beliefs

Consider a Bayesian binary decision-making problem in star networks, where local agents make selfish decisions independently, and a fusion agent makes a final decision based on aggregated decisions and its own private signal. In particular, we assume all agents have private beliefs for the true prior probability, based on which they perform Bayesian decision making. We focus on the Bayes risk of the fusion agent and counterintuitively find that incorrect beliefs could achieve a smaller risk than that when agents know the true prior. It is of independent interest for sociotechnical system design that the optimal beliefs of local agents resemble human probability reweighting models from cumulative prospect theory. We also consider asymptotic characterization of the optimal beliefs and fusion agent's risk in the number of local agents. We find that the optimal risk of the fusion agent converges to zero exponentially fast as the number of local agents grows. Furthermore, having an identical constant belief is asymptotically optimal in the sense of the risk exponent. For additive Gaussian noise, the optimal belief turns out to be a simple function of only error costs and the risk exponent can be explicitly characterized.

cs.IT↗

arXiv2021

Optimizing the Energy Efficiency of Unreliable Memories for Quantized Kalman Filtering

This paper presents a quantized Kalman filter implemented using unreliable memories. We consider that both the quantization and the unreliable memories introduce errors in the computations, and develop an error propagation model that takes into account these two sources of errors. In addition to providing updated Kalman filter equations, the proposed error model accurately predicts the covariance of the estimation error and gives a relation between the performance of the filter and its energy consumption, depending on the noise level in the memories. Then, since memories are responsible for a large part of the energy consumption of embedded systems, optimization methods are introduced so as to minimize the memory energy consumption under a desired estimation performance of the filter. The first method computes the optimal energy levels allocated to each memory bank individually, and the second one optimizes the energy allocation per groups of memory banks. Simulations show a close match between the theoretical analysis and experimental results. Furthermore, they demonstrate an important reduction in energy consumption of more than 50%.

cs.IT↗

arXiv2021

Limits of Detecting Extraterrestrial Civilizations

The search for extraterrestrial intelligence (SETI) is a scientific endeavor which struggles with unique issues -- a strong indeterminacy in what data to look for and when to do so. This has led to attempts at finding both fundamental limits of the communication between extraterrestrial intelligence and human civilizations, as well as benchmarks so as to predict what kinds of signals we might most expect. Previous work has been formulated in terms of the information-theoretic task of communication, but we instead argue it should be viewed as a detection problem, specifically one-shot (asymmetric) hypothesis testing. With this new interpretation, we develop fundamental limits as well as provide simple examples of how to use this framework to analyze and benchmark different possible signals from extraterrestrial civilizations. We show that electromagnetic signaling for detection requires much less power than for communication, that detection as a function of power can be non-linear, and that much of the analysis in this framework may be addressed using computationally efficient optimization problems, thereby demonstrating tools for further inquiry.

cs.IT↗

arXiv2021

Autoequivariant Network Search via Group Decomposition

Recent works show that group equivariance as an inductive bias improves neural network performance for both classification and generation. However, designing group-equivariant neural networks is challenging when the group of interest is large and is unknown. Moreover, inducing equivariance can significantly reduce the number of independent parameters in a network with fixed feature size, affecting its overall performance. We address these problems by proving a new group-theoretic result in the context of equivariant neural networks that shows that a network is equivariant to a large group if and only if it is equivariant to smaller groups from which it is constructed. Using this result, we design a novel fast group equivariant construction algorithm, and a deep Q-learning-based search algorithm in a reduced search space, yielding what we call autoequivariant networks (AENs). AENs find the right balance between equivariance and network size when tested on new benchmark datasets, G-MNIST and G-Fashion-MNIST, obtained via group transformations on MNIST and Fashion-MNIST respectively that we release. Extending these results to group convolutional neural networks, where we optimize between equivariances, augmentations, and network sizes, we find group equivariance to be the most dominating factor in all high-performing GCNNs on several datasets like CIFAR10, SVHN, RotMNIST, ASL, EMNIST, and KMNIST.

cs.LG↗

arXiv2021

Evaluating State-of-the-Art Classification Models Against Bayes Optimality

Evaluating the inherent difficulty of a given data-driven classification problem is important for establishing absolute benchmarks and evaluating progress in the field. To this end, a natural quantity to consider is the \emph{Bayes error}, which measures the optimal classification error theoretically achievable for a given data distribution. While generally an intractable quantity, we show that we can compute the exact Bayes error of generative models learned using normalizing flows. Our technique relies on a fundamental result, which states that the Bayes error is invariant under invertible transformation. Therefore, we can compute the exact Bayes error of the learned flow models by computing it for Gaussian base distributions, which can be done efficiently using Holmes-Diaconis-Ross integration. Moreover, we show that by varying the temperature of the learned flow models, we can generate synthetic datasets that closely resemble standard benchmark datasets, but with almost any desired Bayes error. We use our approach to conduct a thorough investigation of state-of-the-art classification models, and find that in some -- but not all -- cases, these models are capable of obtaining accuracy very near optimal. Finally, we use our method to evaluate the intrinsic "hardness" of standard benchmark datasets, and classes within those datasets.

stat.ML↗

arXiv2021

GAEA: Graph Augmentation for Equitable Access via Reinforcement Learning

Disparate access to resources by different subpopulations is a prevalent issue in societal and sociotechnical networks. For example, urban infrastructure networks may enable certain racial groups to more easily access resources such as high-quality schools, grocery stores, and polling places. Similarly, social networks within universities and organizations may enable certain groups to more easily access people with valuable information or influence. Here we introduce a new class of problems, Graph Augmentation for Equitable Access (GAEA), to enhance equity in networked systems by editing graph edges under budget constraints. We prove such problems are NP-hard, and cannot be approximated within a factor of $(1-\tfrac{1}{3e})$. We develop a principled, sample- and time- efficient Markov Reward Process (MRP)-based mechanism design framework for GAEA. Our algorithm outperforms baselines on a diverse set of synthetic graphs. We further demonstrate the method on real-world networks, by merging public census, school, and transportation datasets for the city of Chicago and applying our algorithm to find human-interpretable edits to the bus network that enhance equitable access to high-quality schools across racial groups. Further experiments on Facebook networks of universities yield sets of new social connections that would increase equitable access to certain attributed nodes across gender groups.

cs.LG↗

arXiv2021

BERTology Meets Biology: Interpreting Attention in Protein Language Models

Transformer architectures have proven to learn useful representations for protein classification and generation tasks. However, these representations present challenges in interpretability. In this work, we demonstrate a set of methods for analyzing protein Transformer models through the lens of attention. We show that attention: (1) captures the folding structure of proteins, connecting amino acids that are far apart in the underlying sequence, but spatially close in the three-dimensional structure, (2) targets binding sites, a key functional component of proteins, and (3) focuses on progressively more complex biophysical properties with increasing layer depth. We find this behavior to be consistent across three Transformer architectures (BERT, ALBERT, XLNet) and two distinct protein datasets. We also present a three-dimensional visualization of the interaction between attention and protein structure. Code for visualization and analysis is available at https://github.com/salesforce/provis.

cs.CL↗

arXiv2021

Wireless Network Coding with Intelligent Reflecting Surfaces

Conventional wireless techniques are becoming inadequate for beyond fifth-generation (5G) networks due to latency and bandwidth considerations. To improve the error performance and throughput of wireless communication systems, we propose physical layer network coding (PNC) in an intelligent reflecting surface (IRS)-assisted environment. We consider an IRS-aided butterfly network, where we propose an algorithm for obtaining the optimal IRS phases. Also, analytic expressions for the bit error rate (BER) are derived. The numerical results demonstrate that the proposed scheme significantly improves the BER performance. For instance, the BER at the relay in the presence of a 32-element IRS is three orders of magnitudes less than that without an IRS.

cs.NI↗

arXiv2021

Adversarial Linear Contextual Bandits with Graph-Structured Side Observations

This paper studies the adversarial graphical contextual bandits, a variant of adversarial multi-armed bandits that leverage two categories of the most common side information: \emph{contexts} and \emph{side observations}. In this setting, a learning agent repeatedly chooses from a set of $K$ actions after being presented with a $d$-dimensional context vector. The agent not only incurs and observes the loss of the chosen action, but also observes the losses of its neighboring actions in the observation structures, which are encoded as a series of feedback graphs. This setting models a variety of applications in social networks, where both contexts and graph-structured side observations are available. Two efficient algorithms are developed based on \texttt{EXP3}. Under mild conditions, our analysis shows that for undirected feedback graphs the first algorithm, \texttt{EXP3-LGC-U}, achieves the regret of order $\mathcal{O}(\sqrt{(K+α(G)d)T\log{K}})$ over the time horizon $T$, where $α(G)$ is the average \emph{independence number} of the feedback graphs. A slightly weaker result is presented for the directed graph setting as well. The second algorithm, \texttt{EXP3-LGC-IX}, is developed for a special class of problems, for which the regret is reduced to $\mathcal{O}(\sqrt{α(G)dT\log{K}\log(KT)})$ for both directed as well as undirected feedback graphs. Numerical tests corroborate the efficiency of proposed algorithms.

cs.LG↗

arXiv2021

The Twelvefold Way of Non-Sequential Lossless Compression

Many information sources are not just sequences of distinguishable symbols but rather have invariances governed by alternative counting paradigms such as permutations, combinations, and partitions. We consider an entire classification of these invariances called the twelvefold way in enumerative combinatorics and develop a method to characterize lossless compression limits. Explicit computations for all twelve settings are carried out for i.i.d. uniform and Bernoulli distributions. Comparisons among settings provide quantitative insight.

cs.IT↗

arXiv2021

Mirostat: A Neural Text Decoding Algorithm that Directly Controls Perplexity

Neural text decoding is important for generating high-quality texts using language models. To generate high-quality text, popular decoding algorithms like top-k, top-p (nucleus), and temperature-based sampling truncate or distort the unreliable low probability tail of the language model. Though these methods generate high-quality text after parameter tuning, they are ad hoc. Not much is known about the control they provide over the statistics of the output, which is important since recent reports show text quality is highest for a specific range of likelihoods. Here, first we provide a theoretical analysis of perplexity in top-k, top-p, and temperature sampling, finding that cross-entropy behaves approximately linearly as a function of p in top-p sampling whereas it is a nonlinear function of k in top-k sampling, under Zipfian statistics. We use this analysis to design a feedback-based adaptive top-k text decoding algorithm called mirostat that generates text (of any length) with a predetermined value of perplexity, and thereby high-quality text without any tuning. Experiments show that for low values of k and p in top-k and top-p sampling, perplexity drops significantly with generated text length, which is also correlated with excessive repetitions in the text (the boredom trap). On the other hand, for large values of k and p, we find that perplexity increases with generated text length, which is correlated with incoherence in the text (confusion trap). Mirostat avoids both traps: experiments show that cross-entropy has a near-linear relation with repetition in generated text. This relation is almost independent of the sampling method but slightly dependent on the model used. Hence, for a given language model, control over perplexity also gives control over repetitions. Experiments with human raters for fluency, coherence, and quality further verify our findings.

cs.CL↗

arXiv2021

Wireless Power Transfer for Future Networks: Signal Processing, Machine Learning, Computing, and Sensing

Wireless power transfer (WPT) is an emerging paradigm that will enable using wireless to its full potential in future networks, not only to convey information but also to deliver energy. Such networks will enable trillions of future low-power devices to sense, compute, connect, and energize anywhere, anytime, and on the move. The design of such future networks brings new challenges and opportunities for signal processing, machine learning, sensing, and computing so as to make the best use of the RF radiations, spectrum, and network infrastructure in providing cost-effective and real-time power supplies to wireless devices and enable wireless-powered applications. In this paper, we first review recent signal processing techniques to make WPT and wireless information and power transfer as efficient as possible. Topics include power amplifier and energy harvester nonlinearities, active and passive beamforming, intelligent reflecting surfaces, receive combining with multi-antenna harvester, modulation, coding, waveform, massive MIMO, channel acquisition, transmit diversity, multi-user power region characterization, coordinated multipoint, and distributed antenna systems. Then, we overview two different design methodologies: the model and optimize approach relying on analytical system models, modern convex optimization, and communication theory, and the learning approach based on data-driven end-to-end learning and physics-based learning. We discuss the pros and cons of each approach, especially when accounting for various nonlinearities in wireless-powered networks, and identify interesting emerging opportunities for the approaches to complement each other. Finally, we identify new emerging wireless technologies where WPT may play a key role -- wireless-powered mobile edge computing and wireless-powered sensing -- arguing WPT, communication, computation, and sensing must be jointly designed.

cs.IT↗

arXiv2020

Mind Your Inflections! Improving NLP for Non-Standard Englishes with Base-Inflection Encoding

Inflectional variation is a common feature of World Englishes such as Colloquial Singapore English and African American Vernacular English. Although comprehension by human readers is usually unimpaired by non-standard inflections, current NLP systems are not yet robust. We propose Base-Inflection Encoding (BITE), a method to tokenize English text by reducing inflected words to their base forms before reinjecting the grammatical information as special symbols. Fine-tuning pretrained NLP models for downstream tasks using our encoding defends against inflectional adversaries while maintaining performance on clean data. Models using BITE generalize better to dialects with non-standard inflections without explicit training and translation models converge faster when trained with BITE. Finally, we show that our encoding improves the vocabulary efficiency of popular data-driven subword tokenizers. Since there has been no prior work on quantitatively evaluating vocabulary efficiency, we propose metrics to do so.

cs.CL↗