Filtern
Erscheinungsjahr
- 2020 (2) (entfernen)
Schlagworte
- Network Analysis (1)
- Network Data (1)
- Netzwerkanalyse (1)
- Netzwerkdaten (1)
The wide accessibility of the Internet and web-based programs enable an increased volume of online interventions for mental health treatment. In contrast to traditional face-to-face therapy, online treatment has the potential to overcome some of the barriers such as improved geographical accessibility, individual time planning, and reduced costs. The availability of clients' treatment data fuels research to analyze the collected data to obtain a better understanding of the relationship among symptoms in mental disorders and derive outcome and symptom predictions. This research leads to predictive models that can be integrated into the online treatment process to assist clinicians and clients. This dissertation discusses different aspects of the development of predictive modeling in online treatment: Categorization of predictive models, data analyses for predictive purposes, and model evaluation. Specifically, the categorization of predictive models and barriers against the uptake of mental health treatment are discussed in the first part of this dissertation. Data analysis and predictive modeling are emphasized in the second part by presenting methods for inference and prediction of mood as well as the prediction of treatment outcome and costs. Prediction of future and current mood can be beneficial in many aspects. Inference of users' mood levels based on unobtrusive measures or diary data can provide crucial information for intervention scheduling. Prediction of future mood can be used to assess clients' response to the treatment and expected treatment outcome. Prediction of the expected treatment costs and outcomes for different treatment types allows simultaneous optimization of these objectives and to increase the cost-effectiveness of the treatment. In the third part, a systematic predictive model evaluation incorporating simulation analyses is demonstrated and a method for model parameter estimation for computationally limited devices is presented. This dissertation aims to overcome the current challenges of predictive model development and its use in online treatment. The development of predictive models for varies data collected in online treatment is demonstrated and how these models can be applied in practice. The derived results contribute to computer science and mental health research with client individual data analysis, the development ofpredictive models, and their statistical evaluation.
Technological development made it possible to store and process data on a scale not imaginable decades ago — a development that also includes network data. A particular characteristic of network data is that, unlike standard data, the objects of interest, called nodes, have relationships to (possibly all) other objects in the network. Collecting empirical data is often complicated and cumbersome, hence, the observed data are typically incomplete and might also contain other types of errors. Because of the interdependent structure of network data, these errors have a severe impact on network analysis methods. This cumulative dissertation is about the impact of erroneous network data on centrality measures, which are methods to assess the position of an object, for example a person, with respect to all other objects in a network. Existing studies have shown that even small errors can substantially alter these positions. The impact of errors on centrality measures is typically quantified using a concept called robustness. The articles included in this dissertation contribute to a better understanding of the robustness of centrality measures in several aspects. It is argued why the robustness needs to be estimated and a new method is proposed. This method allows researchers to estimate the robustness of a centrality measure in a specific network and can be used as a basis for decision making. The relationship between network properties and the robustness of centrality measures is analyzed. Experimental and analytical approaches show that centrality measures are often more robust in networks with a larger average degree. The study of the impact of non-random errors on the robustness suggests that centrality measures are often more robust if missing nodes are more likely to belong to the same community compared to missingness completely at random. For the development of imputation procedures based on machine learning techniques, a process for the evaluation of node embedding methods is proposed.