IA Scholar Query: Computationally Tractable Probabilistic Modeling of Boolean Operators.
https://scholar.archive.org/
Internet Archive Scholar query results feedeninfo@archive.orgThu, 08 Sep 2022 00:00:00 GMTfatcat-scholarhttps://scholar.archive.org/help1440ReX: A Framework for Generating Local Explanations to Recurrent Neural Networks
https://scholar.archive.org/work/auvpquzt4bdtrda5lxuj7fqooy
We propose a general framework to adapt various local explanation techniques to recurrent neural networks. In particular, our explanations add temporal information, which expand explanations generated from existing techniques to cover data points that have different lengths compared to the original input data point. Our approach is general as it only modifies the perturbation model and feature representation of existing techniques without touching their core algorithms. We have instantiated our approach on LIME and Anchors. Our empirical evaluation shows that it effectively improves the usefulness of explanations generated by these two techniques on a sentiment analysis network and an anomaly detection network.Junhao Liu, Xin Zhangwork_auvpquzt4bdtrda5lxuj7fqooyThu, 08 Sep 2022 00:00:00 GMTDepth-efficient proofs of quantumness
https://scholar.archive.org/work/egtwdz5oi5cszme4dopnvm6zre
A proof of quantumness is a type of challenge-response protocol in which a classical verifier can efficiently certify the quantum advantage of an untrusted prover. That is, a quantum prover can correctly answer the verifier's challenges and be accepted, while any polynomial-time classical prover will be rejected with high probability, based on plausible computational assumptions. To answer the verifier's challenges, existing proofs of quantumness typically require the quantum prover to perform a combination of polynomial-size quantum circuits and measurements. In this paper, we give two proof of quantumness constructions in which the prover need only perform constant-depth quantum circuits (and measurements) together with log-depth classical computation. Our first construction is a generic compiler that allows us to translate all existing proofs of quantumness into constant quantum depth versions. Our second construction is based around the learning with rounding problem, and yields circuits with shorter depth and requiring fewer qubits than the generic construction. In addition, the second construction also has some robustness against noise.Zhenning Liu, Alexandru Gheorghiuwork_egtwdz5oi5cszme4dopnvm6zreWed, 07 Sep 2022 00:00:00 GMTDimensional Modeling of Emotions in Text with Appraisal Theories: Corpus Creation, Annotation Reliability, and Prediction
https://scholar.archive.org/work/24asghcvwzd6zm2zeyazvnl5qq
The most prominent tasks in emotion analysis are to assign emotions to texts and to understand how emotions manifest in language. An observation for NLP is that emotions can be communicated implicitly by referring to events, appealing to an empathetic, intersubjective understanding of events, even without explicitly mentioning an emotion name. In psychology, the class of emotion theories known as appraisal theories aims at explaining the link between events and emotions. Appraisals can be formalized as variables that measure a cognitive evaluation by people living through an event that they consider relevant. They include the assessment if an event is novel, if the person considers themselves to be responsible, if it is in line with the own goals, and many others. Such appraisals explain which emotions are developed based on an event, e.g., that a novel situation can induce surprise or one with uncertain consequences could evoke fear. We analyze the suitability of appraisal theories for emotion analysis in text with the goal of understanding if appraisal concepts can reliably be reconstructed by annotators, if they can be predicted by text classifiers, and if appraisal concepts help to identify emotion categories. To achieve that, we compile a corpus by asking people to textually describe events that triggered particular emotions and to disclose their appraisals. Then, we ask readers to reconstruct emotions and appraisals from the text. This setup allows us to measure if emotions and appraisals can be recovered purely from text and provides a human baseline. Our comparison of text classification methods to human annotators shows that both can reliably detect emotions and appraisals with similar performance. Therefore, appraisals constitute an alternative computational emotion analysis paradigm and further improve the categorization of emotions in text with joint models.Enrica Troiano and Laura Oberländer and Roman Klingerwork_24asghcvwzd6zm2zeyazvnl5qqWed, 07 Sep 2022 00:00:00 GMTSafe End-to-end Learning-based Robot Autonomy via Integrated Perception, Planning, and Control
https://scholar.archive.org/work/czetxbsik5favafpkxjvfapvgy
Trustworthy robots must be able to complete tasks reliably while obeying safety constraints. While traditional methods for constrained motion planning and optimal control can achieve this if the environment is accurately modeled and the task is unambiguous, future robots will be deployed in unstructured settings with poorly-understood or inaccurate dynamics, observation models, and task specifications. Thus, to plan and perform control, robots will invariably need data to learn and refine their understanding of their environments and tasks. Though machine learning provides a means to obtain perception and dynamics models from data, blindly trusting these potentially-unreliable models when planning can cause unsafe and unpredictable behavior at runtime. To this end, this dissertation is motivated by the following questions: (1) To refine their understanding of the desired task, how can robots learn components of a constrained motion planner (e.g., constraints, task specifications) in a data-efficient manner? and (2) How can robots quantify and remain robust to the inevitable uncertainty and error in learned components within the broader perception-planning-control autonomy loop in order to provide system-level guarantees on safety and task completion at runtime? To address the first question, we propose methods that use successful human demonstrations to learn unknown constraints and task specifications. The crux of this problem relies on learning what not to do (i.e., behavior violating the unknown constraints or specifications) from only successful examples. We make the insight that the demonstrations' approximate optimality implicitly defines what the robot should not do, and that this information can be extracted by simulating lower-cost trajectories and by using the Karush-Kuhn-Tucker (KKT) optimality conditions. These strong optimality priors make our method highly data-efficient. We use these methods to learn a broad class of constraints, including nonconvex obstacle constraints, and linear temporal logic f [...]Glen Chou, University, Mywork_czetxbsik5favafpkxjvfapvgyTue, 06 Sep 2022 00:00:00 GMTMixing time bounds for edge flipping on regular graphs
https://scholar.archive.org/work/ryssw2gourcq7gnbiokg66bshu
The edge flipping is a non-reversible Markov chain on a given connected graph, which is defined by Chung and Graham in [CG12]. In the same paper, its eigenvalues and stationary distributions for some classes of graphs are identified. We further study its spectral properties to show a lower bound for the rate of convergence in the case of regular graphs. Moreover, we show that a cutoff occurs at 1/4 n logn for the edge flipping on the complete graph by a coupling argument.Yunus Emre Demirci, Ümit Işlak, Alperen Yaşar Özdemirwork_ryssw2gourcq7gnbiokg66bshuTue, 06 Sep 2022 00:00:00 GMTMixing time bounds for edge flipping on regular graphs
https://scholar.archive.org/work/2nxclxcganb6jk743xchd4h3ya
The edge flipping is a non-reversible Markov chain on a given connected graph, which is defined by Chung and Graham. In the same paper, its eigenvalues and stationary distributions for some classes of graphs are identified. We further study its spectral properties to show a lower bound for the rate of convergence in the case of regular graphs. Moreover, we show that a cutoff occurs at 1/4 n log n for the edge flipping on the complete graph by a coupling argumentYunus Emre Demirci, Ümi̇t Işlak, Alperen Özdemirwork_2nxclxcganb6jk743xchd4h3yaSat, 03 Sep 2022 00:00:00 GMTPrivacy-preserving Data Sharing on Vertically Partitioned Data
https://scholar.archive.org/work/zl6a2kundnh35fsmpwr4kcsbxq
In this work, we introduce a differentially private method for generating synthetic data from vertically partitioned data, i.e., where data of the same individuals is distributed across multiple data holders or parties. We present a differentially privacy stochastic gradient descent (DP-SGD) algorithm to train a mixture model over such partitioned data using variational inference. We modify a secure multiparty computation (MPC) framework to combine MPC with differential privacy (DP), in order to use differentially private MPC effectively to learn a probabilistic generative model under DP on such vertically partitioned data. Assuming the mixture components contain no dependencies across different parties, the objective function can be factorized into a sum of products of the contributions calculated by the parties. Finally, MPC is used to compute the aggregate between the different contributions. Moreover, we rigorously define the privacy guarantees with respect to the different players in the system. To demonstrate the accuracy of our method, we run our algorithm on the Adult dataset from the UCI machine learning repository, where we obtain comparable results to the non-partitioned case.Razane Tajeddine, Joonas Jälkö, Samuel Kaski, Antti Honkelawork_zl6a2kundnh35fsmpwr4kcsbxqFri, 02 Sep 2022 00:00:00 GMTA Modern Primer on Processing in Memory
https://scholar.archive.org/work/uhujxt7ctzfk5ivo3a36okxayy
Modern computing systems are overwhelmingly designed to move data to computation. This design choice goes directly against at least three key trends in computing that cause performance, scalability and energy bottlenecks: (1) data access is a key bottleneck as many important applications are increasingly data-intensive, and memory bandwidth and energy do not scale well, (2) energy consumption is a key limiter in almost all computing platforms, especially server and mobile systems, (3) data movement, especially off-chip to on-chip, is very expensive in terms of bandwidth, energy and latency, much more so than computation. These trends are especially severely-felt in the data-intensive server and energy-constrained mobile systems of today. At the same time, conventional memory technology is facing many technology scaling challenges in terms of reliability, energy, and performance. As a result, memory system architects are open to organizing memory in different ways and making it more intelligent, at the expense of higher cost. The emergence of 3D-stacked memory plus logic, the adoption of error correcting codes inside the latest DRAM chips, proliferation of different main memory standards and chips, specialized for different purposes (e.g., graphics, low-power, high bandwidth, low latency), and the necessity of designing new solutions to serious reliability and security issues, such as the RowHammer phenomenon, are an evidence of this trend. This chapter discusses recent research that aims to practically enable computation close to data, an approach we call processing-in-memory (PIM). PIM places computation mechanisms in or near where the data is stored (i.e., inside the memory chips, in the logic layer of 3D-stacked memory, or in the memory controllers), so that data movement between the computation units and memory is reduced or eliminated.Onur Mutlu, Saugata Ghose, Juan Gómez-Luna, Rachata Ausavarungnirunwork_uhujxt7ctzfk5ivo3a36okxayyWed, 31 Aug 2022 00:00:00 GMTMC^2: Rigorous and Efficient Directed Greybox Fuzzing
https://scholar.archive.org/work/7ctki3pa4rhi3pa7dxbhwyi77u
Directed greybox fuzzing is a popular technique for targeted software testing that seeks to find inputs that reach a set of target sites in a program. Most existing directed greybox fuzzers do not provide any theoretical analysis of their performance or optimality. In this paper, we introduce a complexity-theoretic framework to pose directed greybox fuzzing as a oracle-guided search problem where some feedback about the input space (e.g., how close an input is to the target sites) is received by querying an oracle. Our framework assumes that each oracle query can return arbitrary content with a large but constant amount of information. Therefore, we use the number of oracle queries required by a fuzzing algorithm to find a target-reaching input as the performance metric. Using our framework, we design a randomized directed greybox fuzzing algorithm that makes a logarithmic (wrt. the number of all possible inputs) number of queries in expectation to find a target-reaching input. We further prove that the number of oracle queries required by our algorithm is optimal, i.e., no fuzzing algorithm can improve (i.e., minimize) the query count by more than a constant factor. We implement our approach in MC^2 and outperform state-of-the-art directed greybox fuzzers on challenging benchmarks (Magma and Fuzzer Test Suite) by up to two orders of magnitude (i.e., 134×) on average. MC^2 also found 15 previously undiscovered bugs that other state-of-the-art directed greybox fuzzers failed to find.Abhishek Shah, Dongdong She, Samanway Sadhu, Krish Singal, Peter Coffman, Suman Janawork_7ctki3pa4rhi3pa7dxbhwyi77uTue, 30 Aug 2022 00:00:00 GMTSum-of-Squares Relaxations for Information Theory and Variational Inference
https://scholar.archive.org/work/siuyu3e2wvevlctglc27kvry54
We consider extensions of the Shannon relative entropy, referred to as f-divergences. Three classical related computational problems are typically associated with these divergences: (a) estimation from moments, (b) computing normalizing integrals, and (c) variational inference in probabilistic models. These problems are related to one another through convex duality, and for all them, there are many applications throughout data science, and we aim for computationally tractable approximation algorithms that preserve properties of the original problem such as potential convexity or monotonicity. In order to achieve this, we derive a sequence of convex relaxations for computing these divergences from non-centered covariance matrices associated with a given feature vector: starting from the typically non-tractable optimal lower-bound, we consider an additional relaxation based on "sums-of-squares", which is is now computable in polynomial time as a semidefinite program, as well as further computationally more efficient relaxations based on spectral information divergences from quantum information theory. For all of the tasks above, beyond proposing new relaxations, we derive tractable algorithms based on augmented Lagrangians and first-order methods, and we present illustrations on multivariate trigonometric polynomials and functions on the Boolean hypercube.Francis Bachwork_siuyu3e2wvevlctglc27kvry54Mon, 29 Aug 2022 00:00:00 GMTMulti-Winner Voting with Approval Preferences
https://scholar.archive.org/work/kawrnipt7ndipjdhvhumywe72e
Multi-winner voting is the process of selecting a fixed-size set of representative candidates based on voters' preferences. It occurs in applications ranging from politics (parliamentary elections) to the design of modern computer applications (collaborative filtering, dynamic Q&A platforms, diversifying search results). All these applications share the problem of identifying a representative subset of alternatives -- and the study of multi-winner voting is the principled analysis of this task. This book provides a thorough and in-depth look at multi-winner voting based on approval preferences. One speaks of approval preferences if voters express their preferences by providing a set of candidates they approve. Approval preferences thus separate candidates in approved and disapproved ones, a simple, binary classification. The corresponding multi-winner voting rules are called approval-based committee (ABC) rules. Due to the simplicity of approval preferences, ABC rules are widely suitable for practical use. Recent years have seen a rising interest in ABC voting. While multi-winner voting has been originally a topic studied by economists and political scientists, a significant share of recent progress has occurred in the field of computational social choice. This discipline is situated in the intersection of artificial intelligence, computer science, economics, and (to a lesser degree) political science, combining insights and methods from these distinct fields. The goal of this book is to present fundamental concepts and results for ABC voting and to discuss the recent advances in computational social choice. The main focus is on axiomatic analysis, algorithmic results, and relevant applications.Martin Lackner, Piotr Skowronwork_kawrnipt7ndipjdhvhumywe72eMon, 29 Aug 2022 00:00:00 GMTDagstuhl Reports, Volume 12, Issue 1, January 2019, Complete Issue
https://scholar.archive.org/work/sz47bglqcjeoha4rv2m7lcgyba
Dagstuhl Reports, Volume 12, Issue 1, January 2019, Complete Issuework_sz47bglqcjeoha4rv2m7lcgybaTue, 23 Aug 2022 00:00:00 GMTDiscovery and density estimation of latent confounders in Bayesian networks with evidence lower bound
https://scholar.archive.org/work/bpayvv2dj5hg3hwwzopti2tgie
Discovering and parameterising latent confounders represent important and challenging problems in causal structure learning and density estimation respectively. In this paper, we focus on both discovering and learning the distribution of latent confounders. This task requires solutions that come from different areas of statistics and machine learning. We combine elements of variational Bayesian methods, expectation-maximisation, hill-climbing search, and structure learning under the assumption of causal insufficiency. We propose two learning strategies; one that maximises model selection accuracy, and another that improves computational efficiency in exchange for minor reductions in accuracy. The former strategy is suitable for small networks and the latter for moderate size networks. Both learning strategies perform well relative to existing solutions.Kiattikun Chobtham, Anthony C. Constantinouwork_bpayvv2dj5hg3hwwzopti2tgieMon, 22 Aug 2022 00:00:00 GMTLIPIcs, Volume 240, COSIT 2022, Complete Volume
https://scholar.archive.org/work/7m7bfxazsra63myunecp6u6qgm
LIPIcs, Volume 240, COSIT 2022, Complete VolumeToru Ishikawa, Sara Irina Fabrikant, Stephan Winterwork_7m7bfxazsra63myunecp6u6qgmMon, 22 Aug 2022 00:00:00 GMTTruth-Table Net: A New Convolutional Architecture Encodable By Design Into SAT Formulas
https://scholar.archive.org/work/q6wqruybanby5m4bawq66jd4ze
With the expanding role of neural networks, the need for complete and sound verification of their property has become critical. In the recent years, it was established that Binary Neural Networks (BNNs) have an equivalent representation in Boolean logic and can be formally analyzed using logical reasoning tools such as SAT solvers. However, to date, only BNNs can be transformed into a SAT formula. In this work, we introduce Truth Table Deep Convolutional Neural Networks (TTnets), a new family of SAT-encodable models featuring for the first time real-valued weights. Furthermore, it admits, by construction, some valuable conversion features including post-tuning and tractability in the robustness verification setting. The latter property leads to a more compact SAT symbolic encoding than BNNs. This enables the use of general SAT solvers, making property verification easier. We demonstrate the value of TTnets regarding the formal robustness property: TTnets outperform the verified accuracy of all BNNs with a comparable computation time. More generally, they represent a relevant trade-off between all known complete verification methods: TTnets achieve high verified accuracy with fast verification time, being complete with no timeouts. We are exploring here a proof of concept of TTnets for a very important application (complete verification of robustness) and we believe this novel real-valued network constitutes a practical response to the rising need for functional formal verification. We postulate that TTnets can apply to various CNN-based architectures and be extended to other properties such as fairness, fault attack and exact rule extraction.Adrien Benamira, Thomas Peyrin, Bryan Hooi Kuen-Yewwork_q6wqruybanby5m4bawq66jd4zeThu, 18 Aug 2022 00:00:00 GMTNeural Set Function Extensions: Learning with Discrete Functions in High Dimensions
https://scholar.archive.org/work/ebvn2u2sq5hhdexlbzmjgikteq
Integrating functions on discrete domains into neural networks is key to developing their capability to reason about discrete objects. But, discrete domains are (1) not naturally amenable to gradient-based optimization, and (2) incompatible with deep learning architectures that rely on representations in high-dimensional vector spaces. In this work, we address both difficulties for set functions, which capture many important discrete problems. First, we develop a framework for extending set functions onto low-dimensional continuous domains, where many extensions are naturally defined. Our framework subsumes many well-known extensions as special cases. Second, to avoid undesirable low-dimensional neural network bottlenecks, we convert low-dimensional extensions into representations in high-dimensional spaces, taking inspiration from the success of semidefinite programs for combinatorial optimization. Empirically, we observe benefits of our extensions for unsupervised neural combinatorial optimization, in particular with high-dimensional representations.Nikolaos Karalias, Joshua Robinson, Andreas Loukas, Stefanie Jegelkawork_ebvn2u2sq5hhdexlbzmjgikteqMon, 08 Aug 2022 00:00:00 GMTMotivating explanations in Bayesian networks using MAP-independence
https://scholar.archive.org/work/fqzrafbkkjetlgqplbv3kukocy
In decision support systems the motivation and justification of the system's diagnosis or classification is crucial for the acceptance of the system by the human user. In Bayesian networks a diagnosis or classification is typically formalized as the computation of the most probable joint value assignment to the hypothesis variables, given the observed values of the evidence variables (generally known as the MAP problem). While solving the MAP problem gives the most probable explanation of the evidence, the computation is a black box as far as the human user is concerned and it does not give additional insights that allow the user to appreciate and accept the decision. For example, a user might want to know to whether an unobserved variable could potentially (upon observation) impact the explanation, or whether it is irrelevant in this aspect. In this paper we introduce a new concept, MAP- independence, which tries to capture this notion of relevance, and explore its role towards a potential justification of an inference to the best explanation. We formalize several computational problems based on this concept and assess their computational complexity.Johan Kwisthoutwork_fqzrafbkkjetlgqplbv3kukocyFri, 05 Aug 2022 00:00:00 GMTAnticoncentration in Ramsey graphs and a proof of the Erdős-McKay conjecture
https://scholar.archive.org/work/vynnd27mynbz5lf2hs5nhwm5em
An n-vertex graph is called C-Ramsey if it has no clique or independent set of size Clog_2 n (i.e., if it has near-optimal Ramsey behavior). In this paper, we study edge-statistics in Ramsey graphs, in particular obtaining very precise control of the distribution of the number of edges in a random vertex subset of a C-Ramsey graph. This brings together two ongoing lines of research: the study of "random-like" properties of Ramsey graphs and the study of small-ball probabilities for low-degree polynomials of independent random variables. The proof proceeds via an "additive structure" dichotomy on the degree sequence, and involves a wide range of different tools from Fourier analysis, random matrix theory, the theory of Boolean functions, probabilistic combinatorics, and low-rank approximation. One of the consequences of our result is the resolution of an old conjecture of Erdős and McKay, for which Erdős offered one of his notorious monetary prizes.Matthew Kwan, Ashwin Sah, Lisa Sauermann, Mehtaab Sawhneywork_vynnd27mynbz5lf2hs5nhwm5emThu, 04 Aug 2022 00:00:00 GMTCovariant influences for finite discrete dynamical systems
https://scholar.archive.org/work/eiljyhuwnzasbbuysfkkeccx5q
We develop a rigorous theory of external influences on finite discrete dynamical systems, going beyond the perturbation paradigm, in that the external influence need not be a small contribution. Indeed, the covariance condition can be stated as follows: if we evolve the dynamical system for n time steps and then we disturb it, it is the same as first disturbing the system with the same influence and then letting the system evolve for n time steps. Applying the powerful machinery of resource theories, we develop a theory of covariant influences both when there is a purely deterministic evolution and when randomness is involved. Subsequently, we provide necessary and sufficient conditions for the transition between states under deterministic covariant influences and necessary conditions in the presence of stochastic covariant influences, predicting which transitions between states are forbidden. Our approach, for the first time, employs the framework of resource theories, borrowed from quantum information theory, to the study of finite discrete dynamical systems. The laws we articulate unify the behaviour of different types of finite discrete dynamical systems, and their mathematical flavour makes them rigorous and checkable.Carlo Maria Scandolo, Gilad Gour, Barry C. Sanderswork_eiljyhuwnzasbbuysfkkeccx5qSat, 30 Jul 2022 00:00:00 GMTTaxonomy of Machine Learning Safety: A Survey and Primer
https://scholar.archive.org/work/q7shifpmlzfrzhzqa6h2zc324m
The open-world deployment of Machine Learning (ML) algorithms in safety-critical applications such as autonomous vehicles needs to address a variety of ML vulnerabilities such as interpretability, verifiability, and performance limitations. Research explores different approaches to improve ML dependability by proposing new models and training techniques to reduce generalization error, achieve domain adaptation, and detect outlier examples and adversarial attacks. However, there is a missing connection between ongoing ML research and well-established safety principles. In this paper, we present a structured and comprehensive review of ML techniques to improve the dependability of ML algorithms in uncontrolled open-world settings. From this review, we propose the Taxonomy of ML Safety that maps state-of-the-art ML techniques to key engineering safety strategies. Our taxonomy of ML safety presents a safety-oriented categorization of ML techniques to provide guidance for improving dependability of the ML design and development. The proposed taxonomy can serve as a safety checklist to aid designers in improving coverage and diversity of safety strategies employed in any given ML system.Sina Mohseni, Zhiding Yu, Chaowei Xiao, Jay Yadawa, Haotao Wang, Zhangyang Wangwork_q7shifpmlzfrzhzqa6h2zc324mWed, 27 Jul 2022 00:00:00 GMT