Differential effects of COVID-19 lockdowns on well-being: interaction between age, gender and chronotype
Shay Oved, M.Sc. candidate in Industrial Engineering
The lecture will be held on Tuesday, March 9, 2021, at 14:00
Via Zoom
https://us02web.zoom.us/j/81989377320?pwd=SFNuSWtkZUdheG1JN0tqMjRWR3NwUT09
Abstract:
The unprecedented restrictions imposed due to the COVID-19 pandemic, including movement control orders and lockdowns, altered our daily habits, and severely affected our well-being and physiology. The effect of these changes is yet to be fully understood. We analysed highly detailed data on 169 participants for 2-6 months, before and during the second COVID-19 lockdown in Israel. Our entire study was conducted during the COVID-19 pandemic, and therefore is the first to decipher the specific effects of the lockdown from the general effects of the pandemic. We extracted 12 well-being indicators from sensory data of smartwatches and from self-reported questionnaires, filled on a daily basis using a designated mobile application. We used a mixed ANOVA model to study the interplay between age, gender, and chorotype on well-being before and after lockdowns. Our analyses indicate that at the population level, lockdowns resulted in significant changes in mood, sleep duration, sport duration, social encounters, resting heart rate, and the number of steps. The lockdown's adverse effects were greater for young early chronotypes who did not increase their sleep duration, reduced activity level and suffered from significantly reduced mood, and for women, who further suffered an increase in stress levels and a greater decline in social encounters. Our findings underscore that while lockdowns severely impacted our well-being and physiology in general, greater damage has been identified in certain subpopulations. Based on the observed effects, special attention should be given to younger people, who are usually not in the focus of social support, and to women.
Bio:
Shay Oved is an M.Sc. student at the department of Industrial Engineering in Tel Aviv University, specializing in Business Analytics. Shay holds a B.Sc. degree in Industrial Engineering from Tel Aviv University. Her research focuses on personalized medicine based on behavioral information collected from smart phones, smart watches, and self-reported daily questionnaires. The research is being supervised by Dr. Erez Shmueli.
Feature Importance in Gradient Boosting Trees with Cross-Validation Feature Selection
Afek Ilay Adler, M.Sc. candidate in industrial engineering
The lecture will be held on Tuesday, March 9, 2021, at 14:30
Via Zoom
https://us02web.zoom.us/j/81989377320?pwd=SFNuSWtkZUdheG1JN0tqMjRWR3NwUT09
Abstract:
Gradient Boosting Machine (GBM) is among the go-to algorithms on tabular data, which produces state-of-theart results in many prediction tasks. Despite its popularity, the GBM framework suffers from a fundamental flaw in its base learners. Specifically, most implementations utilize decision trees that are typically biased towards categorical variables with large cardinalities. The effect of this bias was extensively studied over the years, mostly in terms of predictive performance. In this work, we extend the scope and study the effect of biased base learners on GBM feature importance (FI) measures. We show that although these implementations demonstrate highly competitive predictive performance, they still, surprisingly, suffer from bias in FI. By utilizing cross-validated (CV) unbiased base learners, we fix this flaw at a relatively low computational cost. We demonstrate the suggested framework in a variety of synthetic and realworld setups, showing a significant improvement in all GBM FI measures while maintaining relatively the same level of prediction accuracy
