Time-series attribution maps with regularized contrastive learning
- URL: http://arxiv.org/abs/2502.12977v1
- Date: Mon, 17 Feb 2025 18:34:25 GMT
- Title: Time-series attribution maps with regularized contrastive learning
- Authors: Steffen Schneider, Rodrigo González Laiz, Anastasiia Filippova, Markus Frey, Mackenzie Weygandt Mathis,
- Abstract summary: gradient-based attribution methods aim to explain decisions of deep learning models but so far lack identifiability guarantees.
Here, we propose a method to generate attribution maps with identifiability guarantees by developing a regularized contrastive learning algorithm trained on time-series data.
We show theoretically that xCEBRA has favorable properties for identifying the Jacobian matrix of the data generating process.
- Score: 1.5503410315996757
- License:
- Abstract: Gradient-based attribution methods aim to explain decisions of deep learning models but so far lack identifiability guarantees. Here, we propose a method to generate attribution maps with identifiability guarantees by developing a regularized contrastive learning algorithm trained on time-series data plus a new attribution method called Inverted Neuron Gradient (collectively named xCEBRA). We show theoretically that xCEBRA has favorable properties for identifying the Jacobian matrix of the data generating process. Empirically, we demonstrate robust approximation of zero vs. non-zero entries in the ground-truth attribution map on synthetic datasets, and significant improvements across previous attribution methods based on feature ablation, Shapley values, and other gradient-based methods. Our work constitutes a first example of identifiable inference of time-series attribution maps and opens avenues to a better understanding of time-series data, such as for neural dynamics and decision-processes within neural networks.
Related papers
- Gradient-based Explanations for Deep Learning Survival Models [0.716879432974126]
We propose a framework for gradient-based explanation methods tailored to survival neural networks.
We analyze the implications of their theoretical assumptions for time-dependent explanations in the survival setting.
We propose effective visualizations incorporating the temporal dimension.
arXiv Detail & Related papers (2025-02-07T14:36:55Z) - Custom DNN using Reward Modulated Inverted STDP Learning for Temporal
Pattern Recognition [0.0]
Temporal spike recognition plays a crucial role in various domains, including anomaly detection, keyword spotting and neuroscience.
This paper presents a novel algorithm for efficient temporal spike pattern recognition on sparse event series data.
arXiv Detail & Related papers (2023-07-15T18:57:27Z) - Generalizing Backpropagation for Gradient-Based Interpretability [103.2998254573497]
We show that the gradient of a model is a special case of a more general formulation using semirings.
This observation allows us to generalize the backpropagation algorithm to efficiently compute other interpretable statistics.
arXiv Detail & Related papers (2023-07-06T15:19:53Z) - TempSAL -- Uncovering Temporal Information for Deep Saliency Prediction [64.63645677568384]
We introduce a novel saliency prediction model that learns to output saliency maps in sequential time intervals.
Our approach locally modulates the saliency predictions by combining the learned temporal maps.
Our code will be publicly available on GitHub.
arXiv Detail & Related papers (2023-01-05T22:10:16Z) - Inducing Gaussian Process Networks [80.40892394020797]
We propose inducing Gaussian process networks (IGN), a simple framework for simultaneously learning the feature space as well as the inducing points.
The inducing points, in particular, are learned directly in the feature space, enabling a seamless representation of complex structured domains.
We report on experimental results for real-world data sets showing that IGNs provide significant advances over state-of-the-art methods.
arXiv Detail & Related papers (2022-04-21T05:27:09Z) - Invariance Learning in Deep Neural Networks with Differentiable Laplace
Approximations [76.82124752950148]
We develop a convenient gradient-based method for selecting the data augmentation.
We use a differentiable Kronecker-factored Laplace approximation to the marginal likelihood as our objective.
arXiv Detail & Related papers (2022-02-22T02:51:11Z) - Convolutional generative adversarial imputation networks for
spatio-temporal missing data in storm surge simulations [86.5302150777089]
Generative Adversarial Imputation Nets (GANs) and GAN-based techniques have attracted attention as unsupervised machine learning methods.
We name our proposed method as Con Conval Generative Adversarial Imputation Nets (Conv-GAIN)
arXiv Detail & Related papers (2021-11-03T03:50:48Z) - Convolutional Shapelet Transform: A new approach for time series
shapelets [1.160208922584163]
We present a new formulation of time series shapelets including the notion of dilation, and a shapelet extraction method based on convolutional kernels.
We show that our method improves on the state-of-the-art for shapelet algorithms, and we show that it can be used to interpret results from convolutional kernels.
arXiv Detail & Related papers (2021-09-28T06:30:42Z) - Learned Factor Graphs for Inference from Stationary Time Sequences [107.63351413549992]
We propose a framework that combines model-based algorithms and data-driven ML tools for stationary time sequences.
neural networks are developed to separately learn specific components of a factor graph describing the distribution of the time sequence.
We present an inference algorithm based on learned stationary factor graphs, which learns to implement the sum-product scheme from labeled data.
arXiv Detail & Related papers (2020-06-05T07:06:19Z) - Learning Representations using Spectral-Biased Random Walks on Graphs [18.369974607582584]
We study how much a probabilistic bias in this process affects the quality of the nodes picked by the process.
We succinctly capture this neighborhood as a probability measure based on the spectrum of the node's neighborhood subgraph represented as a normalized laplacian matrix.
We empirically evaluate our approach against several state-of-the-art node embedding techniques on a wide variety of real-world datasets.
arXiv Detail & Related papers (2020-05-19T20:42:43Z) - Generalization of Change-Point Detection in Time Series Data Based on
Direct Density Ratio Estimation [1.929039244357139]
We show how existing algorithms can be generalized using various binary classification and regression models.
The algorithms are tested on several synthetic and real-world datasets.
arXiv Detail & Related papers (2020-01-17T15:45:38Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.