Semi-supervised feature selection for graph classification

Kong, Xiangnan; Yu, Philip S.

doi:10.1145/1835804.1835905

Cited by 99 publications

(92 citation statements)

References 17 publications

(16 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Though diverse in terms of underlying approaches, most algorithms [11,30] generally assume a database consisting of "positive" and "negative" graphs, and aim at extracting a set significant subgraphs that are frequently present in one class but absent in the other, which subsequently are used as new (binary) features to train classifiers. These approaches, recently being extended to semi-supervised setting [16], uncertain graphs [15], or multiple side-view [4] which we have adapted for our empirical comparison. Another related work is the one developed in [31] that directly addresses the progressive data but in the non-network context.…”

Section: Related Workmentioning

confidence: 99%

Subnetwork Mining with Spatial and Temporal Smoothness

Dang

You

Singh

et al. 2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

In many real-world applications, data is represented in the form of networks with structures and attributes changing over time. The dynamic changes not only happen at nodes/edges, forming local subnetwork processes, but also eventually influence global states of networks. The need to understand what these local network processes are, how they evolve and consequently govern the progression of global network states has become increasingly important. In this paper, we explore these questions and develop a novel algorithm for mining a succinct set of subnetworks that are predictive and evolve along with the progression of global network states. Our algorithm is designed in the framework of logistic regression that fits a model for multi-states of network samples. Its objective function considers both the spatial network topology and temporal smooth transition between adjacent global network states, and we show that its global optimum solution can be achieved via steepest descent. Extensive experimental analysis on both synthetic and real world datasets demonstrates the effectiveness of our algorithm against competing methods, not only in the prediction accuracy but also in terms of domain relevance of the discovered subnetworks.

show abstract

Section: Related Workmentioning

confidence: 99%

Subnetwork Mining with Spatial and Temporal Smoothness

Dang

You

Singh

et al. 2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…In summary, the embedding process is to utilize confidence weight information to help find informative subgraphs as features to represent graphs. Similar assumptions have also been used in the previous work which handles graphs with labeled and unlabeled graphs [11]. Based on the above constraints, we derive an evaluation criterion for J (g,ŵ) as follows:…”

Section: Alternating Optimization Frameworkmentioning

confidence: 99%

Multi-Graph Learning with Positive and Unlabeled Bags

Wu¹,

Hong²,

Pan³

et al. 2014

Proceedings of the 2014 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

In this paper, we formulate a new multi-graph learning task with only positive and unlabeled bags, where labels are only available for bags but not for individual graphs inside the bag. This problem setting raises significant challenges because bag-of-graph setting does not have features to directly represent graph data, and no negative bags exits for deriving discriminative classification models. To solve the challenge, we propose a puMGL learning framework which relies on two iteratively combined processes for multigraph learning: (1) deriving features to represent graphs for learning; and (2) deriving discriminative models with only positive and unlabeled graph bags. For the former, we derive a subgraph scoring criterion to select a set of informative subgraphs to convert each graph into a feature space. To handle unlabeled bags, we assign a weight value to each bag and use the adjusted weight values to select most promising unlabeled bags as negative bags. A margin graph pool (MGP), which contains some representative graphs from positive bags and identified negative bags, is used for selecting subgraphs and training graph classifiers. The iterative subgraph scoring, bag weight updating, and MGP based graph classification forms a closed loop to find optimal subgraphs and most suitable unlabeled bags for multi-graph learning. Experiments and comparisons on real-world multigraph data demonstrate the algorithm performance.

show abstract

“…Transactional data mining aims at discriminating whole graphs based on the occurrence of subgraph patterns (cf. [8] and references therein). The present work instead seeks descriptions of node tuples in terms of their relations.…”

Section: Related Workmentioning

confidence: 99%

Towards Finding Relational Redescriptions

Galbrun

Kimmig

2012

Discovery Science

View full text Add to dashboard Cite

Abstract. This paper introduces relational redescription mining, that is, the task of finding two structurally different patterns that describe nearly the same set of object tuples in a relational dataset. By extending redescription mining beyond propositional and real-valued attributes, it provides a powerful tool to match different relational descriptions of the same concept. As a first step towards solving this general task, we introduce an efficient algorithm that mines one description of a given binary concept. A set of graph patterns is built from frequent path patterns connecting example pairs. Experiments in the domain of explaining kinship terms show that this approach can produce complex descriptions that match explanations by domain experts, while being much faster than a direct relational query mining approach.

show abstract

Semi-supervised feature selection for graph classification

Cited by 99 publications

References 17 publications

Subnetwork Mining with Spatial and Temporal Smoothness

Subnetwork Mining with Spatial and Temporal Smoothness

Multi-Graph Learning with Positive and Unlabeled Bags

Towards Finding Relational Redescriptions

Contact Info

Product

Resources

About