Comparison of a knowledge-driven approach with a data-driven approach to modeling bedtime procrastination
|has title::Comparison of a knowledge-driven approach with a data-driven approach to modeling bedtime procrastination|
|Master:||project within::Cognitive Science|
|Student name:||student name::Erik van den Boogaard|
|Start||start date:=1 February 2017|
|End||end date:=31 July 2017|
|Second supervisor:||Bart Kamphorst|
Bedtime procrastination – failing to go to bed at the intended time, while no external circumstances prevent one from doing so – typically resulting in a lack of sleep, significantly affects health and well-being. From literature we know that bedtime procrastination is influenced by many factors such as personality traits, psychological and physiological factors. My research focuses on comparing a knowledge-driven approach with a data-driven approach to modeling bedtime procrastination. To be able to compare both approaches I will collect empirical data from participants using surveys and sensor data (i.e. Fitbit). To validate a knowledge-driven approach (model of known factors and influences) I will apply parameter tuning to a temporal-causal network model. To predict the amount of bedtime procrastination with a data-driven approach I will apply a machine learning algorithm to identity the most relevant – bedtime procrastination classifying – features. When comparing both approaches I may identify some new features (e.g. from sensor data) that contribute to our knowledge about bedtime procrastination and about potential (new) types of (personalized) interventions.