Transductive Structured Classification through Constrained Min-Cuts
نویسندگان
چکیده
We extend the Blum and Chawla (2001) graph min-cut algorithm to structured problems. This extension can alternatively be viewed as a joint inference method over a set of training and test instances where parts of the instances interact through a prespecified associative network. The method has has an efficient approximation through a linear-programming relaxation. On small training data sets, the method achieves up to 34.8% relative error reduction.
منابع مشابه
Transductive Support Vector Machines
In contrast to learning a general prediction rule, V. Vapnik proposed the transductive learning setting where predictions are made only at a fixed number of known test points. This allows the learning algorithm to exploit the location of the test points, making it a particular type of semi-supervised learning problem. Transductive support vector machines (TSVMs) implement the idea of transducti...
متن کاملTRANS-SC: a Transductive Structural Classifier
The goal of transductive inference (or transduction) is to mine both labeled and unlabeled data within the same learning step and output a classification of the unlabeled examples with few errors as possible. We propose a new transductive classifier, named TRANS-SC, that works in a transductive setting and resorts to a principled probabilistic classification in multi-relational data mining to d...
متن کاملA Feature Space View of Spectral Clustering
The transductive SVM is a semi-supervised learning algorithm that searches for a large margin hyperplane in feature space. By withholding the training labels and adding a constraint that favors balanced clusters, it can be turned into a clustering algorithm. The Normalized Cuts clustering algorithm of Shi and Malik, although originally presented as spectral relaxation of a graph cut problem, ca...
متن کاملRandomized Contractions for Multiobjective Minimum Cuts
We show that Karger’s randomized contraction method [7] can be adapted to multiobjective global minimum cut problems with a constant number of edge or node budget constraints to give efficient algorithms. For global minimum cuts with a single edge-budget constraint, our extension of the randomized contraction method has running time Õ(n3) in an n-node graph improving upon the best-known randomi...
متن کاملMusic Classification by Transductive Learning Using Bipartite Heterogeneous Networks
The popularization of music distribution in electronic format has increased the amount of music with incomplete metadata. The incompleteness of data can hamper some important tasks, such as music and artist recommendation. In this scenario, transductive classification can be used to classify the whole dataset considering just few labeled instances. Usually transductive classification is perform...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2007