Manifold Based Data Refinement for Biological Analysis
- 4 Downloads
This work presents the study into a new manifold method for dimension reduction in digital biological analysis. Extracting features from experiments for multiclass classification task using machine learning is challenging due to different resource populations and various biological sub domains. In data training with a large number of features and samples, errors in classification can occur if efficient feature detection method is not pursued. The aim of the paper is to make clear why some subsets of training samples and features are more appropriate than others. We used Bayesian reasoning under multivariate analysis of learning process to validate and then decrease the number of features used in both training and testing. During training, the number of samples is also reduced by suitability assessment. The method have been designed for rapid and scalable learning by combining selection of features and filtering training samples. Further the article includes experiments of the method with SVM classification model and performance evaluation for digital biological analysis.
- 3.Ma, Y., Fu, Y.: Manifold Learning Theory and Applications. CRC Press Taylor & Francis Group, Boca Raton (2012)Google Scholar
- 6.Xue, H., Song, Y., Xu, H.M.: Multiple indefinite kernel learning for feature selection. In: Sierra, C. (ed.) Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI 2017), pp. 3210-3216. AAAI Press (2017)Google Scholar
- 8.Agbehadji, I.E., Millham, R., Fong, S.J., Yang, H.: Kestrel-Based Search Algorithm (KSA) for parameter tuning unto long short term memory (LSTM) network for feature selection in classification of high-dimensional bioinformatics datasets. In: 2018 Federated Conference on Computer Science and Information Systems (FedCSIS), Poznan, pp. 15-20 (2018)Google Scholar
- 10.Christos, B., Woodruff, D.P.: Optimal CUR matrix decompositions. In: STOC 2014 Proceedings of the Forty-Sixth Annual ACM Symposium on Theory of Computing (2014)Google Scholar
- 16.ISO 5725-1:1994: Accuracy (trueness and precision) of measurement methods and results - part 1: general principles and definitions (1994)Google Scholar