An Empirical Analysis of Backward Compatibility in Machine Learning
Systems
- URL: http://arxiv.org/abs/2008.04572v1
- Date: Tue, 11 Aug 2020 08:10:58 GMT
- Title: An Empirical Analysis of Backward Compatibility in Machine Learning
Systems
- Authors: Megha Srivastava, Besmira Nushi, Ece Kamar, Shital Shah, Eric Horvitz
- Abstract summary: We consider how updates, intended to improve ML models, can introduce new errors that can significantly affect downstream systems and users.
For example, updates in models used in cloud-based classification services, such as image recognition, can cause unexpected erroneous behavior.
- Score: 47.04803977692586
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: In many applications of machine learning (ML), updates are performed with the
goal of enhancing model performance. However, current practices for updating
models rely solely on isolated, aggregate performance analyses, overlooking
important dependencies, expectations, and needs in real-world deployments. We
consider how updates, intended to improve ML models, can introduce new errors
that can significantly affect downstream systems and users. For example,
updates in models used in cloud-based classification services, such as image
recognition, can cause unexpected erroneous behavior in systems that make calls
to the services. Prior work has shown the importance of "backward
compatibility" for maintaining human trust. We study challenges with backward
compatibility across different ML architectures and datasets, focusing on
common settings including data shifts with structured noise and ML employed in
inferential pipelines. Our results show that (i) compatibility issues arise
even without data shift due to optimization stochasticity, (ii) training on
large-scale noisy datasets often results in significant decreases in backward
compatibility even when model accuracy increases, and (iii) distributions of
incompatible points align with noise bias, motivating the need for
compatibility aware de-noising and robustness methods.
Related papers
- Root Causing Prediction Anomalies Using Explainable AI [3.970146574042422]
We present a novel application of explainable AI (XAI) for root-causing performance degradation in machine learning models.
A single feature corruption can cause cascading feature, label and concept drifts.
We have successfully applied this technique to improve the reliability of models used in personalized advertising.
arXiv Detail & Related papers (2024-03-04T19:38:50Z) - Robustness-Congruent Adversarial Training for Secure Machine Learning
Model Updates [13.911586916369108]
We show that misclassifications in machine-learning models can affect robustness to adversarial examples.
We propose a technique, named robustness-congruent adversarial training, to address this issue.
We show that our algorithm and, more generally, learning with non-regression constraints, provides a theoretically-grounded framework to train consistent estimators.
arXiv Detail & Related papers (2024-02-27T10:37:13Z) - Towards Continually Learning Application Performance Models [1.2278517240988065]
Machine learning-based performance models are increasingly being used to build critical job scheduling and application optimization decisions.
Traditionally, these models assume that data distribution does not change as more samples are collected over time.
We develop continually learning performance models that account for the distribution drift, alleviate catastrophic forgetting, and improve generalizability.
arXiv Detail & Related papers (2023-10-25T20:48:46Z) - Robustness and Generalization Performance of Deep Learning Models on
Cyber-Physical Systems: A Comparative Study [71.84852429039881]
Investigation focuses on the models' ability to handle a range of perturbations, such as sensor faults and noise.
We test the generalization and transfer learning capabilities of these models by exposing them to out-of-distribution (OOD) samples.
arXiv Detail & Related papers (2023-06-13T12:43:59Z) - Robustness, Evaluation and Adaptation of Machine Learning Models in the
Wild [4.304803366354879]
We study causes of impaired robustness to domain shifts and present algorithms for training domain robust models.
A key source of model brittleness is due to domain overfitting, which our new training algorithms suppress and instead encourage domain-general hypotheses.
arXiv Detail & Related papers (2023-03-05T21:41:16Z) - How robust are pre-trained models to distribution shift? [82.08946007821184]
We show how spurious correlations affect the performance of popular self-supervised learning (SSL) and auto-encoder based models (AE)
We develop a novel evaluation scheme with the linear head trained on out-of-distribution (OOD) data, to isolate the performance of the pre-trained models from a potential bias of the linear head used for evaluation.
arXiv Detail & Related papers (2022-06-17T16:18:28Z) - Switchable Representation Learning Framework with Self-compatibility [50.48336074436792]
We propose a Switchable representation learning Framework with Self-Compatibility (SFSC)
SFSC generates a series of compatible sub-models with different capacities through one training process.
SFSC achieves state-of-the-art performance on the evaluated datasets.
arXiv Detail & Related papers (2022-06-16T16:46:32Z) - AI Total: Analyzing Security ML Models with Imperfect Data in Production [2.629585075202626]
Development of new machine learning models is typically done on manually curated data sets.
We develop a web-based visualization system that allows the users to quickly gather headline performance numbers.
It also enables the users to immediately observe the root cause of an issue when something goes wrong.
arXiv Detail & Related papers (2021-10-13T20:56:05Z) - Bridging the Gap Between Clean Data Training and Real-World Inference
for Spoken Language Understanding [76.89426311082927]
Existing models are trained on clean data, which causes a textitgap between clean data training and real-world inference.
We propose a method from the perspective of domain adaptation, by which both high- and low-quality samples are embedding into similar vector space.
Experiments on the widely-used dataset, Snips, and large scale in-house dataset (10 million training examples) demonstrate that this method not only outperforms the baseline models on real-world (noisy) corpus but also enhances the robustness, that is, it produces high-quality results under a noisy environment.
arXiv Detail & Related papers (2021-04-13T17:54:33Z) - On Robustness and Transferability of Convolutional Neural Networks [147.71743081671508]
Modern deep convolutional networks (CNNs) are often criticized for not generalizing under distributional shifts.
We study the interplay between out-of-distribution and transfer performance of modern image classification CNNs for the first time.
We find that increasing both the training set and model sizes significantly improve the distributional shift robustness.
arXiv Detail & Related papers (2020-07-16T18:39:04Z)
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.