Auto^6ML Library

Introduction

Auto^6ML is a open-source library for machine learning automation. It is based entirely on jittor, offering high performance and faster speeds. The package supports algorithms based on SLeM(Simulating Learning Methodology ) and some popular meta-learning algorithms.

Our library is divided by methods, which include:

• Data Automation methods
• Network Automation methods
• Loss Automation methods
• Algorithm Automation methods

Supported Methods

The currently supported algorithms include:

popular meta-learning algorithms

Data Automation methods[Code]

• L2W- Learning to Reweight Examples for Robust Deep Learning [ICML 2018] [Code]

Algorithm based on SLeM

Data Automation methods[Code]

• MWNet- Learning an Explicit Mapping For Sample Weighting [NeurIPS 2019] [Code]
• PMWNet- A Probabilistic Formulation for Meta-Weight-Net [TNNLS 2021] [Code]
• CMWNet- CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning [TPAMI 2023] [Code]
• DAC-MR-WNet- sample weighting for robust deep learning [submitted] [Code]
• OS3M- Learning a Trustworthy Sample Selection Strategy for Open-Set Weakly-Supervised Learning [[submitted]] [Code]
• MSLC- label corrector for noisy labels learning [AAAI 2021] [Code]
• CMWNet-SL- an adaptive robust algorithm for sample selection and label correction [NSR 2023] [Code]

Network Automation[Code]

• DPIR- Plug-and-Play Image Restoration With Deep Denoiser Prior [TPAMI 2022] [Code]
• L2AC- Imbalanced Semi-supervised Learning with Bias Adaptive Classifier [ICLR 2023] [Code]
• MTA- Meta Transition Adaptation for Robust Deep Learning with Noisy Labels [submitted] [Code]

Loss Automation[Code]

• HWNet4ACHN- Learning an Explicit Weighting Scheme for Adapting Complex HSI Noise [CVPR 2021] [Code]
• HWNet4HID- A Hyper-weight Network for Hyperspectral Image Denoising [submitted] [Code]
• NARL- Improve noise tolerance of robust loss via noise-awareness [TNNLS 2024] [Code]

Algorithm Automation[Code]

• MLR- Improve noise tolerance of robust loss via noise-awareness [TPAMI 2023] [Code]
• RG- Understanding the Generalization of Bilevel Programming in Hyperparameter Optimization: A Tale of Bias-Variance Decomposition [[submitted]] [Code]

Related

Survey

[1] Jun Shu, Zongben Xu, Deyu Meng. Small sample learning in big data era. 2018.

Framework and Theory

[1] Jun Shu, Deyu Meng, Zongben Xu. Learning an explicit hyperparameter prediction policy conditioned on tasks. JMLR, 2023.