This is my Course Project for Introduction to Data Science (DSA-210 formerly CS-210) at Sabancı Üniversitesi, Fall 2024-2025. This project is an analysis of my main physical and online activity of this semester.
Ever since the pandemic, I have had a major focus and productivity issue; downright crippling. In the years since my attention span and ability to persevere through hard tasks had gone to shreds, I have made small but cumulative efforts towards becoming the productive, ambitious and ever-achieving man I have wanted to become. Two such changes were reducing my social media usage and increasing my physical activity. As this semester is the culmination of all of efforts so far, I want to do an analysis of how I stand against the pressures of university now, i.e: how physically active and online-ically deactive am I and do approaching deadlines and loads of classes cause any fluctations in there? Only one way to find out.
I will be doing a cross on my youtube activity, FitBit recordings, Location history as well as self-annotated important dates in the early weeks of this semester. Data is collected from:
- Google takeout (Youtube, FitBit and location history)
- Self-annotations of semester deadlines and notable moments.
I want to observe the correlations between my activity, physical and online, and my school work (assignments and class hours) to see if certain days of the week or assignments/tests have influence over my activity.
As some basic preliminary guesses as to the findings of my analyses:
- If I have more class hours on a day I will have less steps (paradoxically) on average because I know I tend to miss out on classes out of stress.
- If I have an assignment/test upcoming I will have higher zone minutes (indicating sports) on days leading up to them because I know I that I use basketball as a means of escapism.
- On days outside of campus I will have more steps on average because I spend more time on my feet while outside.
- On nights before tests I will have a higher Youtube watch time because I chronically procrastinate studying by watching Youtube videos
The data from Google takeout is in multiple different file types (json, csv, etc) and so the first steps will be to start begin data visualisation. I also plan on using an AI model to analyse the genre of the Youtube videos watched to see if there is a pattern to the kinds of videos I watch at certain specific times and semester situations. After lots of visualisations of my activity I will begin finding correlations and exploring whether my hypotheses stand true and making further adjacent observations.