Boosting with structure information in the functional space

Fei, Hongliang; Huan, Jun

doi:10.1145/1835804.1835886

Cited by 33 publications

(47 citation statements)

References 44 publications

(71 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Filter approaches are independent of any specific learning algorithms [22,2], while wrapper approaches involve learning algorithms as part of the evaluation procedure [39,23]. Some representative wrapping methods include Sequential Floating Selection (SFS) [39], Sequential Forward Floating Selection (SFFS) [32] and sparse logistic regression based methods [27,29,14,34].…”

Section: Metaheuristics For Feature Selectionmentioning

confidence: 99%

Feature selection for high-dimensional classification using a competitive swarm optimizer

2016

View full text Add to dashboard Cite

When solving many machine learning problems such as classification, there exists a large number of input features. However, not all features are relevant for solving the problem, and sometimes, including irrelevant features may deteriorate the learning performance. Therefore, it is essential to select the most relevant features, which is known as feature selection. Many feature selection algorithms have been developed, including evolutionary algorithms or particle swarm optimization (PSO) algorithms, to find a subset of the most important features for accomplishing a particular machine learning task. However, the traditional PSO does not perform well for large scale optimization problems, which degrades the effectiveness of PSO for feature selection when the number of features dramatically increases. In this paper, we propose to use a very recent PSO variant, known as competitive swarm optimizer (CSO) that was dedicated to large-scale optimization, for solving high-dimensional feature selection problems. In addition, the CSO, which was originally developed for continuous optimization, is adapted to performing feature selection that can be considered as a combinatorial optimization problem. An archive technique is also introduced to reduce computational cost. Experiments on six benchmark datasets demonstrate that compared to the canonical PSO based and a state-of-the-art PSO variant for feature selection, the proposed CSO-based

show abstract

Section: Metaheuristics For Feature Selectionmentioning

confidence: 99%

Feature selection for high-dimensional classification using a competitive swarm optimizer

2016

View full text Add to dashboard Cite

show abstract

“…For subgraph feature based methods, the major goal is to identify significant subgraphs which can be used as signature for different classes [7], [29], [5], [9], [30]. By using subgraph features selected from the graph set, one can easily transfer graphs into a vector space so existing machine learning methods can be applied for classification [7], [29].…”

Section: Related Workmentioning

confidence: 99%

“…After obtaining subgraph features, one can also employ Boosting algorithms for graph classification [34], [5], [9], [35]. In [9], the authors proposed to boost the subgraph decision stumps from frequent subgraphs, which means they first need to provide a minimum support to mine a set of frequent subgraphs and then utilize the function space knowledge for boosting.…”

Section: Related Workmentioning

confidence: 99%

“…Different from traditional data which are represented in deterministic feature space by using an instance-feature representation, structural data are not represented by using attribute vectors, but by graphs with dependency relationships between objects. Because graphs do not have features immediately available for learning, this challenge has recently motivated a number of works using different designs for graph classification, which either learn some global similarities between graphs measured by graph kernels and graph embedding [1], [4], or select some informative subgraphs as features to differentiate graphs in different categories [5], [6], [7], [8], [9], [3], [10], [11]. However, all existing methods for graph classification may suffer two fundamental issues, i.e., ineffective for cost-sensitive classification and inefficient for large graphs, as stated as follows:…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CogBoost: Boosting for Fast Cost-Sensitive Graph Classification

Pan

Zhu

2015

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

Abstract-Graph classification has drawn great interests in recent years due to the increasing number of applications involving objects with complex structural relationships. To date, all existing graph classification algorithms assume, explicitly or implicitly, that misclassifying instances in different classes incurs an equal amount of cost/risk, which is often not the case in real-life applications (where misclassifying a certain class of samples, such as diseased patients, is subject to more expensive costs than others). Although cost-sensitive learning has been extensively studied, all methods are based on data with instance-feature representation. Graphs, however, do not have features available for learning and the possible feature space of graph data are likely infinite and needs to be carefully explored in order to favor classes with a higher cost. Furthermore, current graph classification models are designed for small graph datasets and can hardly scale to the increase of the training set size. In this paper, we propose, CogBoost, a fast cost-sensitive graph classification algorithm, where the goal is to minimize the misclassification costs (instead of the errors) and to achieve fast learning speed for large scale graph datasets. To minimize the misclassification costs, CogBoost iteratively selects the most discriminative subgraph by considering costs of different classes, and then solves a linear programming problem in each iteration by using Bayes decision rule based optimal loss function. In addition, a cutting plane algorithm is derived to speed up the solving of linear programs for fast learning on large graph datasets. Experiments and comparisons on real-world large graph datasets demonstrate the effectiveness and the efficiency of our algorithm.

show abstract

“…This is an extension of the l 2 norm regularized Laplacian on a single vector in [12]. With this addition, we obtain the following optimization problem:…”

Section: Graph Guided Joint Sparse Pcamentioning

confidence: 99%

A Family of Joint Sparse PCA Algorithms for Anomaly Localization in Network Data Streams

Jiang

Fei

Huan

2013

IEEE Trans. Knowl. Data Eng.

Self Cite

View full text Add to dashboard Cite

Determining anomalies in data streams that are collected and transformed from various types of networks has recently attracted significant research interest.Principal Component Analysis (PCA) is arguably the most widely applied unsupervised anomaly detection technique for networked data streams due to its simplicity and efficiency. However, none of existing PCA based approaches addresses the problem of identifying the sources that contribute most to the observed anomaly, or anomaly localization. In this paper, we first proposed a novel joint sparse PCA method to perform anomaly detection and localization for network data streams. Our key observation is that we can detect anomalies and localize anomalous sources by identifying a low dimensional abnormal subspace that captures the abnormal behavior of data. To better capture the sources of anomalies, we incorporated the structure of the network stream data in our anomaly localization framework. Also, an extended version of PCA, multidimensional KLE, was introduced to stabilize the localization performance. We performed comprehensive experimental studies on four real-world data sets from different application domains and compared our proposed techniques with several state-of-the-arts. Our experimental studies demonstrate the utility of the proposed methods.

show abstract

Boosting with structure information in the functional space

Cited by 33 publications

References 44 publications

Feature selection for high-dimensional classification using a competitive swarm optimizer

Feature selection for high-dimensional classification using a competitive swarm optimizer

CogBoost: Boosting for Fast Cost-Sensitive Graph Classification

A Family of Joint Sparse PCA Algorithms for Anomaly Localization in Network Data Streams

Contact Info

Product

Resources

About