Your Mind, Quantified
summary
Built an iOS/Apple Watch app called Nearing iOS/Apple Watch app that keeps data on-device and is free and just launched on the App Store; exported the raw logs to a Python notebook and cleaned a July–September window to extract features: weekly depression score, Apple Watch stress, PCA of activities and emotions, and four semantic clusters from mental‑health notes. Key findings: a suspicious near‑perfect correlation between activity PCA and depression over three months, higher depression tracks higher variability in emotional valence and cumulative negative emotion, and cumulative stress over the prior three weeks (not momentary stress) correlates with higher depression.
AI-written summary · about
chapters (8)
- 0:00Personal journey to Nearing
- 1:53From raw data to health insights
- 2:13Data exploration and cleaning
- 3:17Defining research questions
- 5:27Identifying correlations
- 9:18Key takeaways and future steps
- 10:47Q&A: Data sufficiency and next steps
- 13:43Q&A: User engagement challenges
related talks
17mKnow Thyself: Introspective Personal Data Mining
43mGetting Sentimental: exploration of iOS sentiment analysis
01h 05mWhat the hell am I doing?
48mBuilding a Meaningful Apple Watch Experience
22mSwift Charts - Beyond the basics
21mExamining the relationship between brain activity and video game performance
45mKeynote: From Hypnotised to Heretic: Immunising Society Against Misinformation.
40mPython For Self-Trackers: How To Become A Data-Driven You