A New Ensemble Method for Detecting Anomalies in Gene Expression Matrices

• DOI: 10.3390/math9080882
• Publication Date: 2021-04-19T02:15:13Z
• Authors: Laura Selicato; Flavia Esposito; Grazia Gargano; Maria Carmela Veg...; Giuseppina Opinto; Gian Maria Zaccaria; Sabino Ciavarella; Attilio Guarini; Nicoletta Del Buono
• Keywords: Identification; Expression (computer science); Hierarchical clustering

One of the main problems in analysis real data is often related to presence anomalies. Namely, anomalous cases can both spoil resulting and contain valuable information at same time. In cases, ability detect these occurrences very important. biomedical field, a correct identification outliers could allow development new biological hypotheses that are not considered when looking experimental data. this work, we address problem detecting gene expression data, focusing on microarray analysis. We propose an ensemble approach for anomalies matrices based use Hierarchical Clustering Robust Principal Component Analysis, which allows us derive novel pseudo-mathematical classification