We review the literature on semi-supervised learning, which is an area in machine learning and more generally, artificial intelligence. There has been a whole spectrum of interesting ideas on how to learn from both labeled and unlabeled data, i.e. semi-supervised learning. This document is a chapter excerpt from the author’s doctoral thesis (Zhu, 2005). However the author plans to update the online version frequently to incorporate the latest development in the field. Please obtain the latest version at http://www.cs.wisc.edu/~jerryzhu/pub/ssl_survey.pdf

In this paper, on the one hand, we aim to give a review on literature dealing with the problem of supervised learning aided by additional unlabeled data. On the other hand, being a part of the author's first year PhD report, the paper serves as a frame to bundle related work by the author as well as numerous suggestions for potential future work. Therefore, this work contains more speculative and partly subjective material than the reader might expect from a literature review. We give a rigorous definition of the problem and relate it to supervised and unsupervised learning. The crucial role of prior knowledge is put forward, and we discuss the important notion of input-dependent regularization. We postulate a number of baseline methods, being algorithms or algorithmic schemes which can more or less straightforwardly be applied to the problem, without the need for genuinely new concepts. However, some of them might serve as basis for a genuine method. In the literature revi...

Automatic classification is one of the basic tasks required in any pattern recognition and human computer interaction application. In this paper we discuss training probabilistic classifiers with labeled and unlabeled data. We provide a new analysis that shows under what conditions unlabeled data can be used in learning to improve classification performance. We also show that if the conditions are violated, using unlabeled data can be detrimental to classification performance. We discuss the implications of this analysis to a specific type of probabilistic classifiers, Bayesian networks, and propose a new structure learning algorithm that can utilize unlabeled data to improve classification. Finally, we show how the resulting algorithms are successfully employed in two applications related to human-computer interaction and pattern recognition; facial expression recognition and face detection.

This paper studies the problem of building text classifiers using positive and unlabeled examples. The key feature of this problem is that there is no negative example for learning. Recently, a few techniques for solving this problem were proposed in the literature. These techniques are based on the same idea, which builds a classifier in two steps. Each existing technique uses a different method for each step. In this paper, we first introduce some new methods for the two steps, and perform a comprehensive evaluation of all possible combinations of methods of the two steps. We then propose a more principled approach to solving the problem based on a biased formulation of SVM, and show experimentally that it is more accurate than the existing techniques. 1.

We present a practical co-training method for bootstrapping statistical parsers using a small amount of manually parsed training material and a much larger pool of raw sentences. Experimental results show that unlabelled sentences can be used to improve the performance of statistical parsers. In addition, we consider the problem of bootstrapping parsers when the manually parsed training material is in a different domain to either the raw sentences or the testing material. We show that bootstrapping continues to be useful, even though no manually produced parses from the target domain are used. 1

In traditional text classification, a classifier is built using labeled training documents of every class. This paper studies a different problem. Given a set P of documents of a particular class (called positive class) and a set U of unlabeled documents that contains documents from class P and also other types of documents (called negative class documents), we want to build a classifier to classify the documents in U into documents from P and documents not from P. The key feature of this problem is that there is no labeled negative document, which makes traditional text classification techniques inapplicable. In this paper, we propose an effective technique to solve the problem. It combines the Rocchio method and the SVM technique for classifier building. Experimental results show that the new method outperforms existing methods significantly. 1

One key difficulty with text classification learning algorithms is that they require many hand-labeled examples to learn accurately. This dissertation demonstrates that supervised learning algorithms that use a small number of labeled examples and many inexpensive unlabeled examples can create high-accuracy text classifiers. By assuming that documents are created by a parametric generative model, Expectation-Maximization (EM) finds local maximum a posteriori models and classifiers from all the data -- labeled and unlabeled. These generative models do not capture all the intricacies of text; however on some domains this technique substantially improves classification accuracy, especially when labeled data are sparse. Two problems arise from this basic approach. First, unlabeled data can hurt performance in domains where the generative modeling assumptions are too strongly violated. In this case the assumptions can be made more representative in two ways: by modeling sub-topic class structure, and by modeling super-topic hierarchical class relationships. By doing so, model probability and classification accuracy come into correspondence, allowing unlabeled data to improve classification performance. The second problem is that even with a representative model, the improvements given by unlabeled data do not sufficiently compensate for a paucity of labeled data. Here, limited labeled data provide EM initializations that lead to low-probability models. Performance can be significantly improved by using active learning to select high-quality initializations, and by using alternatives to EM that avoid low-probability local maxima.

Abstract not found

This paper analyzes the performance of semisupervised learning of mixture models. We show that unlabeled data can lead to an increase in classification error even in situations where additional labeled data would decrease classification error. We present a mathematical analysis of this "degradation" phenomenon and show that it is due to the fact that bias may be adversely affected by unlabeled data. We discuss the impact of these theoretical results to practical situations.

Traditionally, text classifiers are built from labeled training examples. Labeling is usually done manually by human experts (or the users), which is a labor intensive and time consuming process. In the past few years, researchers investigated various forms of semi-supervised learning to reduce the burden of manual labeling. In this paper, we propose a different approach. Instead of labeling a set of documents, the proposed method labels a set of representative words for each class. It then uses these words to extract a set of documents for each class from a set of unlabeled documents to form the initial training set. The EM algorithm is then applied to build the classifier. The key issue of the approach is how to obtain a set of representative words for each class. One way is to ask the user to provide them, which is difficult because the user usually can only give a few words (which are insufficient for accurate learning). We propose a method to solve the problem. It combines clustering and feature selection. The technique can effectively rank the words in the unlabeled set according to their importance. The user then selects/labels some words from the ranked list for each class. This process requires less effort than providing words with no help or manual labeling of documents. Our results show that the new method is highly effective and promising.