iDECODe: In-Distribution Equivariance for Conformal Out-of-Distribution Detection

Kaur, Ramneet; Jha, Susmit; Roy, Anirban; Park, Sang‐Don; Dobriban, Edgar; Sokolsky, Oleg; Lee, Insup

doi:10.1609/aaai.v36i7.20670

Cited by 15 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We show that SDEs alleviate the saturation effect faced by attribution methods and empirically demonstrate this. In future efforts, one can explore how such neural SDEs can lead to more robust confidence metrics (Jha et al 2019) and enhance out-of-distribution detection algorithms (Kaur et al 2022).…”

Section: Discussionmentioning

confidence: 99%

Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations

Jha

Ewetz

Velasquez

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

Neural SDEs with Brownian motion as noise lead to smoother attributions than traditional ResNets. Various attribution methods such as saliency maps, integrated gradients, DeepSHAP and DeepLIFT have been shown to be more robust for neural SDEs than for ResNets using the recently proposed sensitivity metric. In this paper, we show that neural SDEs with adaptive attribution-driven noise lead to even more robust attributions and smaller sensitivity metrics than traditional neural SDEs with Brownian motion as noise. In particular, attribution-driven shaping of noise leads to 6.7%, 6.9% and 19.4% smaller sensitivity metric for integrated gradients computed on three discrete approximations of neural SDEs with standard Brownian motion noise: stochastic ResNet-50, WideResNet-101 and ResNeXt-101 models respectively. The neural SDE model with adaptive attribution-driven noise leads to 25.7% and 4.8% improvement in the SIC metric over traditional ResNets and Neural SDEs with Brownian motion as noise. To the best of our knowledge, we are the first to propose the use of attributions for shaping the noise injected in neural SDEs, and demonstrate that this process leads to more robust attributions than traditional neural SDEs with standard Brownian motion as noise.

show abstract

Section: Discussionmentioning

confidence: 99%

Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations

Jha

Ewetz

Velasquez

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

show abstract

“…Split conformal prediction has been extended to provide conditional probabilistic guarantees (Vovk 2012), to handle distribution shifts (Tibshirani et al 2019;Fannjiang et al 2022), and to allow for quantile regression (Romano, Patterson, and Candes 2019). Further, split conformal prediction has been used to construct probably approximately correct prediction sets for machine learning models (Park et al 2020;Angelopoulos et al 2022), to perform out-ofdistribution detection (Kaur et al 2022(Kaur et al , 2023, to guarantee safety in autonomous systems (Luo et al 2022), and to quantify uncertainty for F1/10 car motion predictions (Tumu et al 2023). Additionally, in (Stutz et al 2022) the authors encode the width of the generated prediction sets directly into the loss function of a neural network during training.…”

Section: Related Workmentioning

confidence: 99%

Conformal Prediction Regions for Time Series Using Linear Complementarity Programming

Cleaveland,

Lee,

Pappas

et al. 2024

AAAI

View full text Add to dashboard Cite

Conformal prediction is a statistical tool for producing prediction regions of machine learning models that are valid with high probability. However, applying conformal prediction to time series data leads to conservative prediction regions. In fact, to obtain prediction regions over T time steps with confidence 1--delta, previous works require that each individual prediction region is valid with confidence 1--delta/T. We propose an optimization-based method for reducing this conservatism to enable long horizon planning and verification when using learning-enabled time series predictors. Instead of considering prediction errors individually at each time step, we consider a parameterized prediction error over multiple time steps. By optimizing the parameters over an additional dataset, we find prediction regions that are not conservative. We show that this problem can be cast as a mixed integer linear complementarity program (MILCP), which we then relax into a linear complementarity program (LCP). Additionally, we prove that the relaxed LP has the same optimal cost as the original MILCP. Finally, we demonstrate the efficacy of our method on case studies using pedestrian trajectory predictors and F16 fighter jet altitude predictors.

show abstract

“…Although originally based on the premise of exchangeable (e.g., independently and identically distributed) training and test data, the framework has since been generalized to handle various forms of distribution shift, including covariate shift ( 4 , 7 ), label shift ( 8 ), arbitrary distribution shifts in an online setting ( 6 ), and test distributions that are nearby the training distribution ( 5 ). Conformal approaches have also been used to detect distribution shift ( 17 – 23 ).…”

Section: Uncertainty Quantification Under Feedback Loopsmentioning

confidence: 99%

Conformal prediction under feedback covariate shift for biomolecular design

Fannjiang

Bates

Angelopoulos

et al. 2022

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

Many applications of machine-learning methods involve an iterative protocol in which data are collected, a model is trained, and then outputs of that model are used to choose what data to consider next. For example, a data-driven approach for designing proteins is to train a regression model to predict the fitness of protein sequences and then use it to propose new sequences believed to exhibit greater fitness than observed in the training data. Since validating designed sequences in the wet laboratory is typically costly, it is important to quantify the uncertainty in the model’s predictions. This is challenging because of a characteristic type of distribution shift between the training and test data that arises in the design setting—one in which the training and test data are statistically dependent, as the latter is chosen based on the former. Consequently, the model’s error on the test data—that is, the designed sequences—has an unknown and possibly complex relationship with its error on the training data. We introduce a method to construct confidence sets for predictions in such settings, which account for the dependence between the training and test data. The confidence sets we construct have finite-sample guarantees that hold for any regression model, even when it is used to choose the test-time input distribution. As a motivating use case, we use real datasets to demonstrate how our method quantifies uncertainty for the predicted fitness of designed proteins and can therefore be used to select design algorithms that achieve acceptable tradeoffs between high predicted fitness and low predictive uncertainty.

show abstract

iDECODe: In-Distribution Equivariance for Conformal Out-of-Distribution Detection

Cited by 15 publications

References 32 publications

Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations

Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations

Conformal Prediction Regions for Time Series Using Linear Complementarity Programming

Conformal prediction under feedback covariate shift for biomolecular design

Contact Info

Product

Resources

About