sstliao

Time-series-analysis-on-US-Industrial-Production

This project focuses on multivariate time series and thus we mainly use the tests and models, such as VAR and VARX, which can deal with multivariate time series.

BayesianGraph

Classification-on-Bank-Marketing-Data-Set-to-predict-client-s-intention-of-a-term-subscription

Before using any models, SMOTE algorithm is used to adjust the imbalanced data. Ensemble methods and SVM are also compared for classification in this data via ROC curve.

Implementation-of-Sparse--Group-Lasso-Methods-on-Gene-Expression-Dataset

To deal with too many predictors in large data set, "sparse-group lasso" is given in group lasso, inferring predictors into different groups and get estimates.

shutingliao