Graduate Student Researcher
• Developed and compared novel supervised machine learning models to assess aspects of cognitive function using features derived from passive smartphone typing behaviors.• Engineered features from smartphone typing and accelerometer metadata relating to diurnal patterns and sleep for use in predictive models of cognition.• Designed unsupervised machine learning method to process accelerometer metadata collected during smartphone typing to compare to patterns in mood and cognition.•… Show more • Developed and compared novel supervised machine learning models to assess aspects of cognitive function using features derived from passive smartphone typing behaviors.• Engineered features from smartphone typing and accelerometer metadata relating to diurnal patterns and sleep for use in predictive models of cognition.• Designed unsupervised machine learning method to process accelerometer metadata collected during smartphone typing to compare to patterns in mood and cognition.• Explored influence of time of day on response inhibition as measured by Go/No-Go task in those with mood disorders using mixed effects models.• Identified individual characteristics in diurnal smartphone orientation through accelerometer metadata for detection of mood changes.• Analyzed smartphone keyboard metadata to determine correlation between typing speed, mood, and trail-making test time using hierarchical mixed-effects models. Show less