idea: using insulin pump data to specify the start times
early thesis submission: no special form, 1 Apr deadline
Progress
next version of event-based machine-tested on days 1 & 2
bug fixes for placement of matching time-series
Todo
add captions to images
investigate how library chooses start time
order matches by similarity distance from left to right
Bob draw visual interface of software design
Bob & Kai EuroVis reviews due 20 Jan
first drat of treemap implementation
check if library has most frequent time-series pattern motif
unfortunately, no. We can split the data by hours and iterate seed events to find the number of matches, therefore the most common seed events. The library doesn't support the discovery of the most common arbitrary pattern.