In this paper, we formulate the problem of early classification of time series data, which is important in some time-sensitive applications such as health-informatics. We introduce a novel concept of MPL (Minimum Prediction Length) and develop ECTS (Early Classification on Time Series), an effective 1-nearest neighbor classification method. ECTS makes early predictions and at the same time retains the accuracy comparable to that of a 1NN classifier using the full-length time series. Our empirical study using benchmark time series data sets shows that ECTS works well on the real data sets where 1NN classification is effective.

In machine learning problems, labeled data are often in short supply. One of the feasible solution for this problem is transfer learning. It can make use of the labeled data from other domain to discriminate those unlabeled data in the target domain. In this paper, we propose a transfer learning framework based on similarity matrix approximation to tackle such problems. Two practical algorithms are proposed, which are the label propagation and the similarity propagation. In these methods, we build a hybrid graph based on all available data. Then the information is transferred cross domains through alternatively constructing the similarity matrix for different part of the graph. Among all related methods, similarity propagation approach can make maximum use of all available similarity information across domains. This leads to more efficient transfer and better learning result. The experiment on real world text mining applications demonstrates the promise and effectiveness of our algorithms.

In this paper, we consider a general problem of semi-supervised preference learning, in which we assume that we have the information of the extreme cases and some ordered constraints, our goal is to learn the unknown preferences of the other places. Taking the potential housing place selection problem as an example, we have many candidate places together with their associated information (e.g., position, environment), and we know some extreme examples (i.e., several places are perfect for building a house, and several places are the worst that cannot build a house there), and we know some partially ordered constraints (i.e., for two places, which place is better), then how can we judge the preference of one potential place whose preference is unknown beforehand? We propose a Bayesian framework based on Gaussian process to tackle this problem, from which we not only solve for the unknown preferences, but also the hyperparameters contained in our model.

Clustering aggregation has emerged as an important extension of the classical clustering problem. It refers to the situation in which a number of different (input) clusterings have been obtained for a particular data set and it is desired to aggregate those clustering results to get a better clustering solution. In this paper, we propose a unified framework to solve the clustering aggregation problem, where the aggregated clustering result is obtained by minimizing the (weighted) sum of the Bregman divergence between it and all the input clusterings. Moreover, under our algorithm framework, we also propose a novel cluster aggregation problem where some must-link and cannot-link constraints are given in addition to the input clusterings. Finally the experimental results on some real world data sets are presented to show the effectiveness of our method.

Manifold alignment has been found to be useful in many areas of machine learning and data mining. In this paper we introduce a novel manifold alignment approach, which differs from semi-supervised alignment and Procrustes alignment in that it does not require predetermining correspondences. Our approach learns a projection that maps data instances (from two different spaces) to a lower dimensional space simultaneously matching the local geometry and preserving the neighborhood relationship within each set. This approach also builds connections between spaces defined by different features and makes direct knowledge transfer possible. The performance of our algorithm is demonstrated and validated in a series of carefully designed experiments in information retrieval and bioinformatics.

Kernel discriminant analysis (KDA) is an effective approach for supervised nonlinear dimensionality reduction. Probabilistic models can be used with KDA to improve its robustness. However, the state of the art of such models could only handle binary class problems, which confines their application in many real world problems. To overcome this limitation, we propose a novel nonparametric probabilistic model based on Gaussian Process for KDA to handle multiclass problems. The model provides a novel Bayesian interpretation for KDA, which allows its parameters to be automatically tuned through the optimization of the marginal loglikelihood of the data. Empirical study demonstrates the efficacy of the proposed model.

This paper addresses the problem of activity and event discovery in multi dimensional time series data by proposing a novel method for locating multi dimensional motifs in time series. While recent work has been done in finding single dimensional and multi dimensional motifs in time series, we address motifs in general case, where the elements of multi dimensional motifs have temporal, length, and frequency variations. The proposed method is validated by synthetic data, and empirical evaluation has been done on several wearable systems that are used by real subjects.

For classification with multiple labels, a common approach is to learn a classifier for each label. With a kernel-based classifier, there are two options to set up kernels: select a specific kernel for each label or the same kernel for all labels.  In this work, we present a unified framework for multi-label multiple kernel learning, in which the above two approaches can be considered as two extreme cases. Moreover, our framework allows the kernels shared partially among multiple labels, enabling flexible degrees of label commonality. We systematically study how the sharing of kernels among multiple labels affects the performance based on extensive experiments on various benchmark data including images and microarray data. Interesting findings concerning efficacy and efficiency are reported.

A common approach to the control problem in partially observable environments is to perform a direct search in policy space, as defined over some set of features of history. In this paper we consider predictive features, whose values are conditional probabilities of future events, given history. Since predictive features provide direct information about the agent's future, they have a number of advantages for control. However, unlike more typical features defined directly over past observations, it is not clear how to maintain the values of predictive features over time. A model could be used, since a model can make any prediction about the future, but in many cases learning a model is infeasible. In this paper we demonstrate that in some cases it is possible to learn to maintain the values of a set of predictive features even when a learning a model is infeasible, and that natural predictive features can be useful for policy-search methods.

Practical data analysis relies on the ability to count observations of objectssuccinctly and efficiently. Unfortunately the space usage of an exact estimator grows with the size of the a priori set from which objects are drawn while the time required to maintain such an estimator grows with the size of the data set. We present static and on-line approximation schemes that avoid these limitations when approximate frequency estimates are acceptable. Our Log-Frequency Sketch extends the approximate counting algorithm of Morris [Morris1978] to estimate frequencies with bounded relative error via a single pass over a data set. It uses constant space per object when the frequencies follow a power law and can be maintained in constant time per observation. We give an (epsilon, delta)-approximation scheme which we verify empirically on a large natural language data set where, for instance, 95 percent of frequencies are estimated with relative error less than 0.25 using fewer than 11 bits per object in the static case and 15 bits per object on-line.