Skip to content
conftrace_
2022 AAAI AAAI 2022

Wind Prediction under Random Data Corruption (Student Abstract)

Abstract

Abstract We study the robustness of ridge regression, lasso regression, and of a neural network, when the training set has been randomly corrupted and in response to this corruption the training-size is reduced in order to remove the corrupted data. While the neural network appears to be the most robust method among these three, nevertheless lasso regression appears to be the method of choice since it suffers less loss both when the full information is available to the learner, as well as when a significant amount of the original training set has been rendered useless because of random data corruption.

Authors

Conner Flansburg, Dimitrios I. Diochnos

Topics

Machine Learning > Core Methods > Regression Machine Learning > Application Areas > Efficient Computing Machine Learning > Learning Types > Regularization Machine Learning > Learning Types > Robustness

Keywords

robust regression ridge regression lasso regression data corruption neural network wind prediction
Download PDF

More from AAAI 2022

All AAAI 2022 papers →
Visual Explanations for Convolutional Neural Networks via Latent Traversal of Generative Adversarial Networks (Student Abstract)
Identifying ATT&CK Tactics in Android Malware Control Flow Graph through Graph Representation Learning and Interpretability (Student Abstract)
Reinforcement Learning Explainability via Model Transforms (Student Abstract)
Knowledge-Enhanced Scene Graph Generation with Multimodal Relation Alignment (Student Abstract)
HuggingMolecules: An Open-Source Library for Transformer-Based Molecular Property Prediction (Student Abstract)
From Video to Images: Contrastive Pretraining for Emotion Recognition from Single Image (Student Abstract)