Missing data is common in clinical research. How these missing valuesare handled has a direct impact on the final study results. Existing medi-cal studies commonly use complete case analysis to...Show moreMissing data is common in clinical research. How these missing valuesare handled has a direct impact on the final study results. Existing medi-cal studies commonly use complete case analysis to remove observationswith missing values, which has the advantage that it is simple and easybut depletes the information in the original dataset, and may result in bi-ased estimated. Multiple imputation (MI) methods are often consideredmore reliable than complete case analysis, missing indicator methods andsingle imputation methods. However, recent research has shown that bycomparing a number of MI methods, particularly where the underlyingassumptions are undermined, some MI methods may cause more bias inmodel estimates than complete case analysis.To study which methods would perform better under which circum-stances, this thesis will perform a simulation study, comparing the resultsof the above-mentioned techniques under certain types of missingness,such as MCAR, MAR and MNAR. Complicated connections like miss-ingness is also correlated with survival time will be considered.The various parameter settings for the simulation study are based ona real case study where about 12% of the observations contain missingvalues for some variables. In addition to the basic MICE, two other multi-ple imputation methods are compared, one with interaction terms betweenthe full variables and the baseline hazard in the imputation model, and theother with a specific substantive model in the iteration. In this thesis, thesubstantive model is Cox model. The simulation studies show that theone with interaction terms is not significantly different from MICE andits improvements are of limited applicability. The one with the specificsubstantive model is more suitable for complex data types and when thereare strong correlations between covariates. Besides, basic MICE also per-forms well in data sets with a high proportion of missing binary covariates, while the missing indicator method produces large bias in many settings, even for full case studies.Show less
Excessive data collection can be expensive to perform and process. Feature selection methods such as penalized regression are useful techniques to exclude less relevant features from a prediction...Show moreExcessive data collection can be expensive to perform and process. Feature selection methods such as penalized regression are useful techniques to exclude less relevant features from a prediction model. However, these methods cannot be directly applied to multi-view data, as it ignores the grouped structure of this type of data and valuable information may be lost. For multi-view data Stacked Penalized Regression (StaPLR) can be used for view selection. A challenge arises when missing data is present, especially when complete views are missing. Using listwise deletion, the most common method to deal with missing values, on these datasets will therefore likely lead to a lot of information loss. In this thesis the effect of mean substitution, single Bayesian regression, multiple Bayesian regression and predictive mean matching on the predictive accuracy and sparsity of StaPLR are compared on different missing data scenarios on an existing dataset containing items from the Mood and Anxiety Symptoms Questionnaire (MASQ). In addition to traditional feature imputation, StaPLR offers the possibility to impute view-level predictions when views are missing completely, which can decrease the computational burden. To evaluate the performance of the missing data methods, 27 missing data scenarios are created with features missing either randomly or within the view structure. Predictive accuracy is measured with AUC, and sparsity by the number of views StaPLR selects. Multiple imputation works well in terms of predictive accuracy, but when quantities of missing values increases they do not work well in terms of sparsity. Mean substitution could potentially be an acceptable alternative to multiple imputation when features are missing randomly. View-level imputation performs similar as feature-level imputation for most methods when views are missing completely, and could therefore be used for these methods to decrease the computational burden. This thesis is the first step in handling missing values with StaPLR. Future research directions include repeating this research on more elaborate multi-view data structures, where the views are collected from different sources, to identify whether these effects are found under different circumstances.Show less
