MetAL: Active Semi-Supervised Learning on Graphs via Meta Learning

The objective of active learning (AL) is to train classification models with less number of labeled instances by selecting only the most informative instances for labeling. The AL algorithms designed for other data types such as images and text do not perform well on graph-structured data. Although a few heuristics-based AL algorithms have been proposed for graphs, a principled approach is lacking. In this paper, we propose MetAL, an AL approach that selects unlabeled instances that directly improve the future performance of a classification model. For a semi-supervised learning problem, we formulate the AL task as a bilevel optimization problem. Based on recent work in meta-learning, we use the meta-gradients to approximate the impact of retraining the model with any unlabeled instance on the model performance. Using multiple graph datasets belonging to different domains, we demonstrate that MetAL efficiently outperforms existing state-of-the-art AL algorithms.

[1]  Andrew McCallum,et al.  Toward Optimal Active Learning through Monte Carlo Estimation of Error Reduction , 2001, ICML 2001.

[2]  Paolo Frasconi,et al.  Bilevel Programming for Hyperparameter Optimization and Meta-Learning , 2018, ICML.

[3]  Stephan Gunnemann,et al.  Adversarial Attacks on Graph Neural Networks via Meta Learning , 2019, ICLR.

[4]  Lise Getoor,et al.  Collective Classification in Network Data , 2008, AI Mag..

[5]  Zoubin Ghahramani,et al.  Deep Bayesian Active Learning with Image Data , 2017, ICML.

[6]  Ruslan Salakhutdinov,et al.  Revisiting Semi-Supervised Learning with Graph Embeddings , 2016, ICML.

[7]  Zoubin Ghahramani,et al.  Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning , 2015, ICML.

[8]  Oriol Vinyals,et al.  Matching Networks for One Shot Learning , 2016, NIPS.

[9]  Dale Schuurmans,et al.  Discriminative Batch Mode Active Learning , 2007, NIPS.

[10]  Anton van den Hengel,et al.  Image-Based Recommendations on Styles and Substitutes , 2015, SIGIR.

[11]  Kilian Q. Weinberger,et al.  Simplifying Graph Convolutional Networks , 2019, ICML.

[12]  John Langford,et al.  Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds , 2019, ICLR.

[13]  Ye Zhang,et al.  Active Discriminative Text Representation Learning , 2016, AAAI.

[14]  Percy Liang,et al.  Understanding Black-box Predictions via Influence Functions , 2017, ICML.

[15]  Yuan Luo,et al.  Graph Convolutional Networks for Text Classification , 2018, AAAI.

[16]  Chelsea Finn,et al.  Active One-shot Learning , 2017, ArXiv.

[17]  Lise Getoor,et al.  Active Learning for Networked Data , 2010, ICML.

[18]  W. R. Thompson.  ON THE LIKELIHOOD THAT ONE UNKNOWN PROBABILITY EXCEEDS ANOTHER IN VIEW OF THE EVIDENCE OF TWO SAMPLES , 1933 .

[19]  Jure Leskovec,et al.  Graph Convolutional Neural Networks for Web-Scale Recommender Systems , 2018, KDD.

[20]  Peter Auer,et al.  Using Confidence Bounds for Exploitation-Exploration Trade-offs , 2003, J. Mach. Learn. Res..

[21]  Richard S. Zemel,et al.  Gated Graph Sequence Neural Networks , 2015, ICLR.

[22]  Patrice Marcotte,et al.  An overview of bilevel optimization , 2007, Ann. Oper. Res..

[23]  Max Welling,et al.  Semi-Supervised Classification with Graph Convolutional Networks , 2016, ICLR.

[24]  Zoubin Ghahramani,et al.  Bayesian Active Learning for Classification and Preference Learning , 2011, ArXiv.

[25]  Jan Kautz,et al.  Hierarchical Subquery Evaluation for Active Learning on a Graph , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[26]  Rasoul Karimi,et al.  Active Learning for Recommender Systems , 2015, KI - Künstliche Intelligenz.

[27]  Jiawei Han,et al.  Towards Active Learning on Graphs: An Error Bound Minimization Approach , 2012, 2012 IEEE 12th International Conference on Data Mining.

[28]  Jiawei Han,et al.  A Variance Minimization Criterion to Active Learning on Graphs , 2012, AISTATS.

[29]  Steven Skiena,et al.  DeepWalk: online learning of social representations , 2014, KDD.

[30]  Philip Bachman,et al.  Learning Algorithms for Active Learning , 2017, ICML.

[31]  Burr Settles,et al.  Active Learning Literature Survey , 2009 .

[32]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[33]  Hong Yang,et al.  Active Discriminative Network Representation Learning , 2018, IJCAI.

[34]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[35]  Kevin Chen-Chuan Chang,et al.  Active Learning for Graph Embedding , 2017, ArXiv.