I don’t know much about machine learning, but wouldn’t it be possible to use machine learning to get a machine to optimize your diet, exercise, sleep patterns, behaviour, etc.? Perhaps it generates a list of proposed daily routines, you follow one and report back some stats about yourself like weight, blood pressure, mood, digit span, etc.. It then takes these and uses them to figure out what parts of what daily routines do what. If it suspects eating cinnamon decreases your blood pressure, it makes you eat cinnamon so you can tell it whether it worked. The algorithm can optimize diet, exercise, mental exercises, even choose what books you read.
Basically, I’m saying why don’t we try something like what Piotr Wozniak did with the SRS algorithms, except instead of optimizing memorization we optimize everything. We do what the people at QS do, except we delegate interpretation of the data to a computer.
Like I said, I don’t know much about machine learning, but even the techniques I /do/ know, evolutionary algorithms and neural nets, seem like they could be used for this, and are certainly worth our time trying.
Sounds like it could work, especially if it uses a database of all users, so that users most similar to you also give an indication of what might or might not work for you.
“I am [demographic and psychological parameters] and would like to [specific goal—mood, weight, memoty, knowledge] in the coming [time period]; what would work best?”
Sounds like an interesting project, I’ll have to think about it.
I think machine learning has potential in badly formalized fields. Surely, diet and exercise are not most well formalized fields, but it looks like there are certain working heuristics out there. What do you mean by “behaviour” btw?
To start thinking about applying machine learning to diet, exercise, sleep patterns and behaviour you should answer the question: what do you want to optimize them for?
It surprises and disappoints me that I haven’t heard of some sort of massive expert program like this being used in healthcare yet. I hope it will come soon, perhaps in the form of a Watson derivative.
I don’t know much about machine learning, but wouldn’t it be possible to use machine learning to get a machine to optimize your diet, exercise, sleep patterns, behaviour, etc.? Perhaps it generates a list of proposed daily routines, you follow one and report back some stats about yourself like weight, blood pressure, mood, digit span, etc.. It then takes these and uses them to figure out what parts of what daily routines do what. If it suspects eating cinnamon decreases your blood pressure, it makes you eat cinnamon so you can tell it whether it worked. The algorithm can optimize diet, exercise, mental exercises, even choose what books you read.
Basically, I’m saying why don’t we try something like what Piotr Wozniak did with the SRS algorithms, except instead of optimizing memorization we optimize everything. We do what the people at QS do, except we delegate interpretation of the data to a computer.
Like I said, I don’t know much about machine learning, but even the techniques I /do/ know, evolutionary algorithms and neural nets, seem like they could be used for this, and are certainly worth our time trying.
Sounds like it could work, especially if it uses a database of all users, so that users most similar to you also give an indication of what might or might not work for you.
“I am [demographic and psychological parameters] and would like to [specific goal—mood, weight, memoty, knowledge] in the coming [time period]; what would work best?”
Sounds like an interesting project, I’ll have to think about it.
I think machine learning has potential in badly formalized fields. Surely, diet and exercise are not most well formalized fields, but it looks like there are certain working heuristics out there. What do you mean by “behaviour” btw?
To start thinking about applying machine learning to diet, exercise, sleep patterns and behaviour you should answer the question: what do you want to optimize them for?
It surprises and disappoints me that I haven’t heard of some sort of massive expert program like this being used in healthcare yet. I hope it will come soon, perhaps in the form of a Watson derivative.