Semi-Supervised Learning with Very Few Labeled Training Examples

In semi-supervised learning, a number of labeled examples are usually required for training an initial weakly useful predictor which is in turn used for exploiting the unlabeled examples. However, in many real-world applications there may exist very few labeled training examples, which makes the weakly useful predictor difficult to generate, and therefore these semisupervised learning methods cannot be applied. This paper proposes a method working under a two-view setting. By taking advantages of the correlations between the views using canonical component analysis, the proposed method can perform semi-supervised learning with only one labeled training example. Experiments and an application to content-based image retrieval validate the effectiveness of the proposed method.

[1]  H. Hotelling.  Relations Between Two Sets of Variates , 1936 .

[2]  David J. Miller,et al.  A Mixture of Experts Classifier with Learning Based on Both Labelled and Unlabelled Data , 1996, NIPS.

[3]  Avrim Blum,et al.  The Bottleneck , 2021, Monopsony Capitalism.

[4]  Nicholas Kushmerick,et al.  Learning to remove Internet advertisements , 1999, AGENTS '99.

[5]  Yan Zhou,et al.  Enhancing Supervised Learning with Unlabeled Data , 2000, ICML.

[6]  Avrim Blum,et al.  Learning from Labeled and Unlabeled Data using Graph Mincuts , 2001, ICML.

[7]  Sanjoy Dasgupta,et al.  PAC Generalization Bounds for Co-training , 2001, NIPS.

[8]  Anoop Sarkar,et al.  Applying Co-Training Methods to Statistical Parsing , 2001, NAACL.

[9]  Michael Fink,et al.  Object Classification from a Single Example Utilizing Class Relevance Metrics , 2004, NIPS.

[10]  Maria-Florina Balcan,et al.  Co-Training and Expansion: Towards Bridging Theory and Practice , 2004, NIPS.

[11]  John Shawe-Taylor,et al.  Canonical Correlation Analysis: An Overview with Application to Learning Methods , 2004, Neural Computation.

[12]  Thomas Hofmann,et al.  Semi-supervised Learning on Directed Graphs , 2004, NIPS.

[13]  Sebastian Thrun,et al.  Text Classification from Labeled and Unlabeled Documents using EM , 2000, Machine Learning.

[14]  Mikhail Belkin,et al.  Semi-Supervised Learning on Riemannian Manifolds , 2004, Machine Learning.

[15]  Nicu Sebe,et al.  A new analysis of the value of unlabeled data in semi-supervised learning for image retrieval , 2004, 2004 IEEE International Conference on Multimedia and Expo (ICME) (IEEE Cat. No.04TH8763).

[16]  Naonori Ueda,et al.  A Hybrid Generative/Discriminative Approach to Semi-Supervised Classifier Design , 2005, AAAI.

[17]  Wei-Ying Ma,et al.  A probabilistic semantic model for image annotation and multi-modal image retrieval , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[18]  Wei-Ying Ma,et al.  A Probabilistic Semantic Model for Image Annotation and Multi-Modal Image Retrieva , 2005, ICCV.

[19]  Xiaojin Zhu,et al.  --1 CONTENTS , 2006 .

[20]  Zhi-Hua Zhou,et al.  Tri-training: exploiting unlabeled data using three classifiers , 2005, IEEE Transactions on Knowledge and Data Engineering.

[21]  Gilles Blanchard,et al.  Pattern Recognition from One Example by Chopping , 2005, NIPS.

[22]  Zhi-Hua Zhou,et al.  Enhancing relevance feedback in image retrieval using unlabeled data , 2006, ACM Trans. Inf. Syst..

[23]  Alexander Zien,et al.  Semi-Supervised Learning , 2006 .

[24]  Chris H. Q. Ding,et al.  PSoL: a positive sample only learning algorithm for finding non-coding RNA genes , 2006, Bioinform..