Changes

=== Goals ===

=== Goals ===

Simply adding colorful plots that Anki does not already have will encourage studying. The goal of teaching is almost as certain because analyzing the data requires delving into the learning process. This goal may actually be the most impactful.  

Simply adding colorful plots that Anki does not already have will encourage studying.  

 

The goal of teaching is almost as easy to guarantee because analyzing the data requires delving into the learning process. This goal may actually be the most impactful.  

Project will advise user on how to optimize their studying by comparing what actually happened with what would have happened if they had done something different, according to a machine learning model. Effects will be illustrated using Partial Dependence Plots and ICE<ref>Visualizing ML Models with LIME · UC Business Analytics R Programming Guide (uc-r.github.io)</ref>. To help user conduct experiments, project will compare one part of time series with another or progress on one set of cards with another. User may have to do some analysis on their own.     

Project will advise user on how to optimize their studying by comparing what actually happened with what would have happened if they had done something different, according to a machine learning model. Effects will be illustrated using Partial Dependence Plots and ICE<ref>Visualizing ML Models with LIME · UC Business Analytics R Programming Guide (uc-r.github.io)</ref>. To help user conduct experiments, project will compare one part of time series with another or progress on one set of cards with another. User may have to do some analysis on their own.     

Any decent skill test will detect when subject is severely sick. For a test to be useful to improving the skill it must detect more subtle patterns. Statistics detects plenty of patterns that are obviously not generated by the same process. Unfortunately, those patterns could easily be artifacts of the analysis or test taking process. Adding the confounders to a machine learning model that spits out the actual test results should help but will likely take lots of time and is not guaranteed to remove everything.      

Any decent skill test will detect when subject is severely sick. For a test to be useful for improving the skill, it must detect more subtle patterns. Statistical tests detect plenty of patterns in target data that are subtle enough and obviously not generated by the same process as the rest of the data. Unfortunately, those patterns could easily be artifacts of the analysis or of the test taking process. Adding the confounders to a machine learning model and then using the error / residuals as the actual test results should help, but is not guaranteed to remove everything.      

Everyone should track their cognitive ability as much as health-conscious people track their heartrate and exercise. Formal cognitive testing takes too much time, relevance to user's productivity goals is often questionable. An alternative that solves both is skill trainers like typing tutors. To generalize to cognition and health the skill test must correlate with other cognitive things or health things. Wozniak of supermemo has found correlation between sleep and flash card skill.     

Everyone should track their cognitive ability as much as health-conscious people track their heartrate and exercise. Formal cognitive testing takes too much time, relevance to user's productivity goals is often questionable. An alternative that solves both is skill trainers like typing tutors. To generalize to cognition and health the skill test must correlate with other cognitive things or health things. Wozniak of supermemo has found correlation between sleep and flash card skill.