arXiv research

Randomly initialized networks can perform as well as pruned networks.

Reliability is a critical consideration to DL-based systems. But the statistical nature of DL makes it quite vulnerable to invalid inputs, i.e., those cases that are not considered in the training phase of a DL model. This paper proposes to perform data sanity check to identify invalid inputs, so as to enhance the reli…

There is great interest in "saliency methods" (also called "attribution methods"), which give "explanations" for a deep net's decision, by assigning a "score" to each feature/pixel in the input. Their design usually involves credit-assignment via the gradient of the output with respect to input. Recently Adebayo et al.…

Differentially private learning on real-world data poses challenges for standard machine learning practice: privacy guarantees are difficult to interpret, hyperparameter tuning on private data reduces the privacy budget, and ad-hoc privacy attacks are often required to test model privacy. We introduce three tools to ma…

Saliency methods can aid understanding of deep neural networks. Recent years have witnessed many improvements to saliency methods, as well as new ways for evaluating them. In this paper, we 1) present a novel region-based attribution method, XRAI, that builds upon integrated gradients (Sundararajan et al. 2017), 2) int…

Study tests if input gradients highlight discriminative features, finds they often fail.

Paper presents a universal baseline for binary prediction models.

FDN improves probabilistic regressors' adaptability to distribution shifts.

Probability density estimation is a classical and well studied problem, but standard density estimation methods have historically lacked the power to model complex and high-dimensional image distributions. More recent generative models leverage the power of neural networks to implicitly learn and represent probability …

The paper introduces sanity tests to detect spurious correlations in AI-guided radiology systems.

Two modified tests improve the reliability of evaluating explanation methods.

Detecting adversarial examples is as hard as classifying them.

The ability to perform effective off-policy learning would revolutionize the process of building better interactive systems, such as search engines and recommendation systems for e-commerce, computational advertising and news. Recent approaches for off-policy evaluation and learning in these settings appear promising. …

Recently, several methods have been proposed to explain the predictions of recurrent neural networks (RNNs), in particular of LSTMs. The goal of these methods is to understand the network's decisions by assigning to each input variable, e.g., a word, a relevance indicating to which extent it contributed to a particular…

Explaining the output of a complicated machine learning model like a deep neural network (DNN) is a central challenge in machine learning. Several proposed local explanation methods address this issue by identifying what dimensions of a single input are most responsible for a DNN's output. The goal of this work is to a…

Saliency methods have emerged as a popular tool to highlight features in an input deemed relevant for the prediction of a learned model. Several saliency methods have been proposed, often guided by visual appeal on image data. In this work, we propose an actionable methodology to evaluate what kinds of explanations a g…

Saliency maps are a popular approach to creating post-hoc explanations of image classifier outputs. These methods produce estimates of the relevance of each pixel to the classification output score, which can be displayed as a saliency map that highlights important pixels. Despite a proliferation of such methods, littl…

New method interprets deep neural networks for better recommendation system understanding.

We aim to create the highest possible quality of treatment-control matches for categorical data in the potential outcomes framework. Matching methods are heavily used in the social sciences due to their interpretability, but most matching methods do not pass basic sanity checks: they fail when irrelevant variables are …

The paper tackles confidence calibration for exploratory machine learning problems.

RayS attack improves hard-label adversarial attacks by reducing query complexity and identifying false robust models.

Proposes a copula-based filter for diabetes risk prediction.

Large dataset released for ITE and UM research.

Study examines XAI methods for ECG analysis to improve model transparency.

In this article, we derive concentration inequalities for the cross-validation estimate of the generalization error for empirical risk minimizers. In the general setting, we prove sanity-check bounds in the spirit of \cite{KR99} \textquotedblleft\textit{bounds showing that the worst-case error of this estimate is not m…

Study on sensor fusion algorithms under high dimensional noise.

Scientific discovery is limited by hypothesis redundancy, and hybrid methods can exploit non-local exploration.

Study reveals biases in ImageNet models are not sufficient for generalization.

Deep learning techniques have proven high accuracy for identifying melanoma in digitised dermoscopic images. A strength is that these methods are not constrained by features that are pre-defined by human semantics. A down-side is that it is difficult to understand the rationale of the model predictions and to identify …

Proposes MEED framework for model interpretation.

LLM forecasting benchmarks suffer from information leakage, which confounds model performance.

We give a definition of an integer-valued function $\sum_i α_i x ^*_i$ derived from arrow diagrams for the ambient isotopy classes of oriented spherical curves. Then, we introduce certain elements of the free $\mathbb{Z}$-module generated by the arrow diagrams with at most $l$ arrows, called relators of Type~($\check{…

Unified analysis of multilabel Fisher discriminants with improved dimensionality and robustness.

Unified analysis of multilabel Fisher discriminants with improved dimensionality and robustness.

This paper deforms complex tori and their mirrors using gerbes.

Statistical model checking for PCTL on MDPs using reinforcement learning.

Research aims to make fact-checking models more transparent.

Time-aware fact-checking improves veracity predictions for time-sensitive claims.

DC-Check helps guide ML development by considering data-centric aspects.

This study optimizes trading strategy parameters using walk-forward techniques and finds robust performance.

Proves SYZ mirror symmetry for del Pezzo and rational elliptic surfaces.

Social networks are getting closer to our real physical world. People share the exact location and time of their check-ins and are influenced by their friends. Modeling the spatio-temporal behavior of users in social networks is of great importance for predicting the future behavior of users, controlling the users' mov…

We describe the infinitesimal moduli space of pairs $(Y, V)$ where $Y$ is a manifold with $G_2$ holonomy, and $V$ is a vector bundle on $Y$ with an instanton connection. These structures arise in connection to the moduli space of heterotic string compactifications on compact and non-compact seven dimensional spaces, e.…

By the SYZ construction, a mirror pair $(X,\check{X})$ of a complex torus $X$ and a mirror partner $\check{X}$ of the complex torus $X$ is described as the special Lagrangian torus fibrations $X \rightarrow B$ and $\check{X} \rightarrow B$ on the same base space $B$. Then, by the SYZ transform, we can construct a simpl…

We propose a neural network approach to price EU call options that significantly outperforms some existing pricing models and comes with guarantees that its predictions are economically reasonable. To achieve this, we introduce a class of gated neural networks that automatically learn to divide-and-conquer the problem …

We prove the following result announced in Todorov and Valov: Any homogeneous, metric $ANR$-continuum is a $V^n_G$-continuum provided $\dim_GX=n\geq 1$ and $\check{H}^n(X;G)\neq 0$, where $G$ is a principal ideal domain. This implies that any homogeneous $n$-dimensional metric $ANR$-continuum with $\check{H}^n(X;G)\neq…

We specify a result of Yokoi \cite{yo} by proving that if $G$ is an abelian group and $X$ is a homogeneous metric $ANR$ compactum with $\dim_GX=n$ and $\check{H}^n(X;G)\neq 0$, then $X$ is an $(n,G)$-bubble. This implies that any such space $X$ has the following properties: $\check{H}^{n-1}(A;G)\neq 0$ for every closed…

Paper checks SSC for matrix factorizations using Gurobi.