Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Machine Learning and Ketosis

github.com

Post reply on HN
Print view View original

31–40 of 309 posts

Re: Machine Learning and Ketosis

#31

Post by benkuhn »

aab0 wrote: post #23
benkuhn wrote: post #15

It's astonishing (and really awesome) that they were able to extract such a strong signal from daily weight swings! This makes me a lot more optimistic about the possibilities of quantified-self-type stuff. Three things I'm really curious about: - How significant are the estimates of lifestyle factors? Do you have p-values? If you bootstrap resample, how much do the rankings change at the extremes? - How much cogniti…

> - How significant are the estimates of lifestyle factors? Do you have p-values? If you bootstrap resample, how much do the rankings change at the extremes? His VW script does do bootstrapping ('--bootstrap 16') but he doesn't report it anywhere I see. https://github.com/JohnLangford/vowpal_wabbit/wiki/using-vw-... seems to not report any sort of p-value or confidence interval which might be derived from the bootstr…

I mean, the uncertainties in the middle are less interesting than the uncertainties at the extremes. I'm sure you can't say anything useful about melon, but it would be interesting to know if (e.g.) "sleep" was consistently that big. He did regularize somewhat (`--l2 1.85201e-08`) but unclear whether that's enough regularization.

Basically I'd love to see some actual diagnostics I guess :)

PS I believe the "relevance" is just the coefficient divided by the biggest coefficient (the `RelScore` column in the printout in the README file).

Re: Machine Learning and Ketosis

#32

Post by aab0 »

benkuhn wrote: post #31
aab0 wrote: post #23

Earlier quoted context omitted.

> - How significant are the estimates of lifestyle factors? Do you have p-values? If you bootstrap resample, how much do the rankings change at the extremes? His VW script does do bootstrapping ('--bootstrap 16') but he doesn't report it anywhere I see. https://github.com/JohnLangford/vowpal_wabbit/wiki/using-vw-... seems to not report any sort of p-value or confidence interval which might be derived from the bootstr…

I mean, the uncertainties in the middle are less interesting than the uncertainties at the extremes. I'm sure you can't say anything useful about melon, but it would be interesting to know if (e.g.) "sleep" was consistently that big. He did regularize somewhat (`--l2 1.85201e-08`) but unclear whether that's enough regularization. Basically I'd love to see some actual diagnostics I guess :) PS I believe the "relevance…

Well, he also disables the holdout set (`--holdout_off ariel.train`), so between that and not reporting the uncertainties and his absolutely tiny dataset with high measurement error and the low prior probability that any of these effects could be that big... Nah. It's just overfitting noise.

I tried to run the makefile, but apparently the version of Vowpal Rabbit that ships on my Ubuntu (7.3) is so outdated that it doesn't support the bootstrap option.

Re: Machine Learning and Ketosis

#33

Post by mrfusion »

philip1209 wrote: post #2

I've been doing ketosis for about 6 months, after having done twice in the past. From my heaviest to today over the last 2 years, I'm down 50 pounds. The data in this post was cool. I found Dr. Peter Addia's analysis to be one of my favorite resources. He's a medical doctor who was an overweight endurance athlete, then began doing ketosis. I appreciated an honest scientific analysis of the benefits and drawbacks of k…

What do you mean about insulin?

Re: Machine Learning and Ketosis

#34

Post by zzleeper »

> The 'stayhome' lifestlye, which fell mostly on weekends, is a red-herring, I simply slept longer when I didn't have to commute to work.

Is it?

First, whatever method you use should already take into account that sleep happens together with stayathome. Even basic regressions take into account that.

Second, staying at home means your eating binges are constrained by what's around you. If it is healthy stuff it might mean weight loss, if it is bread and chips the opposite

Re: Machine Learning and Ketosis

#35

Post by djsumdog »

escoz wrote: post #28

If anybody is reading this and curious about Ketosis, I'd recommend Taubes book ( https://www.amazon.com/Why-We-Get-Fat-About/dp/0307474259 ). It's a great review of scientific studies done over the years. I read that 4 years ago, spent another 4 months reading the listed studies, convinced myself it was a good plan, and lost 40 pounds with no exercises. I still do LCHF after all these years, and likely will never go…

I was told about Keto diets about three years back by a friend of mine. There's a lot of misinformation about food and health in the US and elsewhere, and for the longest time I thought low-carb diets were stupid.

But after trying it and reading a lot of the research, I was pretty amazing. I went from close to 70kg down to 58kg over the course of the next year (although I've come up to around 62 ~ 63kg stable).

I like how this person mentions in the Github readme that it's not for everyone and "listen to your body." Some people have bodies that do well with high carb intakes. Every body type is different. With all that being said though, a huge issue with being overweight is education. The food industry wants you to buy fast food, pizzas and things that are cheap to create with easy base ingredients (sugar, starch, corn, wheat, etc.)

Sadly, the only way for me to realize this was to experiment on myself, as did this person (although with totally insane amounts of metrics). Kudos!

Re: Machine Learning and Ketosis

#36

Post by frankus »

Totally anecdotal, but I want to second the recommendation of a longer fasting period.

I switched to a strict-but-not-religious "no food between 7 pm and 11 am" system (with exceptions for weekends and social occasions).

Within a few months I was down 15 pounds (~182 to around ~167) and had shed 4 inches off my waist (~34 to ~30). I'm about 5'10" and 41 years old.

It's definitely helped with physical activity (mostly parkour/free running) and I look better. It's also more convenient than what I was doing before, since I don't have to cook breakfast.

The only negative side effect (possibly unrelated) has been that I need a much cooler sleeping environment to be comfortable.

The only thing I would add is that I'm starting to (upside-down) plateau at around 165 and (what my scale says is) 20% body fat. I would love to lose another 5-10 pounds but it'll probably be a slow process.

Re: Machine Learning and Ketosis

#37

Post by PerfectElement »

thisisananth wrote: post #16

This matches my experience of weight loss using high fat diet. I started it after hearing Sarah Hallberg's Tedx talk. https://www.youtube.com/watch?v=da1vvigy5tQ&feature=youtu.be For a vegetarian, starting and continuing the high fat low carb diet is difficult. Are there any resources for more vegetarian recipes for high fat low carb food? The article was very impressive. I liked the graphs and presentation.

https://www.reddit.com/r/veganketo/

Re: Machine Learning and Ketosis

#38

Post by apineda »

I just started my keto diet about 2 weeks ago and I feel great. Did have one rough night of "keto flu" but then I started supplementing with minerals and more water. I'm super low carb right now. Funny thing is to get (about 4 weeks ago) I would get a double big mac or a double quarter pounder but no drink and no fries and that was it till breakfast. That was fun.

Re: Machine Learning and Ketosis

#39

Post by hoodwink »

tunnuz wrote: post #3

This is all good if your goal is weight loss. However weight loss doesn't necessarily means higher fitness. Glycogen is fundamental if you do sports, and exercise is a major ingredient in getting fit. If I read correctly, exercise was not a big part of your experiment, how would you suggest modifying the experiment to accommodate one's exercising needs? On a side note, I reached a similar conclusion on the role of "c…

Glycogen is necessary for high intensity exercise, but you are mistaken that carbs are necessary for glycogen. When you are keto-adapted and eat low carb, two things happen that most people overlook. First, you get enough "trace carbs" in your diet that you have glycogen in your muscles for exercise. Second, your body can synthesize glycogen!

I've eaten LCHF for 10 years and regularly exercise with CrossFit (arguably the highest intensity workout most people will do) and yoga intermixed with athletic activities like hiking, skiing, and surfing.

My suggestion is to properly keto-adapt over 3 to 6 weeks before concluding that you can't exercise without eating carbs.

Re: Machine Learning and Ketosis

#40

Post by hoodwink »

As a long-term ketogenic eater (10 years), here are my top simple tips. Unfortunately they were not gleaned using machine learning.

1. Watch your protein. Most people when first going keto will eat too much protein and not enough fat. Protein has an insulinogenic effect when eaten in quantity. Keep protein below 8 oz per meal. Don't be afraid to eat more fat.

2. Avoid cheese. Yes, it's technically low carb, but it repeatedly throws me and my girlfriend off (also a low carber).

3. Avoid nuts. Yes, like cheese, nuts are delicious. But they're a slippery slope. Life will be easier if you avoid them.

Post reply on HN