TY - GEN
T1 - A learning framework to improve unsupervised gene network inference
AU - Turki, Turki
AU - Bassett, William
AU - Wang, Jason T.L.
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2016.
PY - 2016
Y1 - 2016
N2 - Network inference through link prediction is an important data mining problem that finds many applications in computational social science and biomedicine. For example, by predicting links, i.e., regulatory relationships, between genes to infer gene regulatory networks (GRNs), computational biologists gain a better understanding of the functional elements and regulatory circuits in cells. Unsupervised methods have been widely used to infer GRNs; however, these methods often create missing and spurious links. In this paper, we propose a learning framework to improve the unsupervised methods. Given a network constructed by an unsupervised method, the proposed framework employs a graph sparsification technique for network sampling and principal component analysis for feature selection to obtain better quality training data, which guides three classifiers to predict and clean the links of the given network. The three classifiers include neural networks, random forests and support vector machines. Experimental results on several datasets demonstrate the good performance of the proposed learning framework and the classifiers used in the framework.
AB - Network inference through link prediction is an important data mining problem that finds many applications in computational social science and biomedicine. For example, by predicting links, i.e., regulatory relationships, between genes to infer gene regulatory networks (GRNs), computational biologists gain a better understanding of the functional elements and regulatory circuits in cells. Unsupervised methods have been widely used to infer GRNs; however, these methods often create missing and spurious links. In this paper, we propose a learning framework to improve the unsupervised methods. Given a network constructed by an unsupervised method, the proposed framework employs a graph sparsification technique for network sampling and principal component analysis for feature selection to obtain better quality training data, which guides three classifiers to predict and clean the links of the given network. The three classifiers include neural networks, random forests and support vector machines. Experimental results on several datasets demonstrate the good performance of the proposed learning framework and the classifiers used in the framework.
KW - Applications in biology and medicine
KW - Feature selection
KW - Graph mining
KW - Network analysis
UR - http://www.scopus.com/inward/record.url?scp=84978969869&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84978969869&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-41920-6_3
DO - 10.1007/978-3-319-41920-6_3
M3 - Conference contribution
AN - SCOPUS:84978969869
SN - 9783319419190
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 28
EP - 42
BT - Machine Learning and Data Mining in Pattern Recognition - 12th International Conference, MLDM 2016, Proceedings
A2 - Perner, Petra
PB - Springer Verlag
T2 - 12th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2016
Y2 - 16 July 2016 through 21 July 2016
ER -