66 reputation
4
bio website
location
age
visits member for 2 years, 1 month
seen Dec 9 '13 at 0:44

Feb
13
awarded  Nice Question
Feb
9
reviewed Excellent Why does the suprailliac skinfold body fat method depend on age?
Feb
9
reviewed Excellent How can you do a BMI test?
Feb
9
reviewed Satisfactory Supercompensation and split workouts
Feb
9
reviewed Excellent What is an ideal body fat% for young lean males?
Feb
9
reviewed Excellent best way to build cardio endurance?
Feb
9
reviewed Excellent Compound, total body exercises minimising quad and ham development
Feb
9
reviewed Satisfactory Quads vs Hamstrings in Squats and Lunges
Feb
9
reviewed Excellent Workout plan for a lazy person
Feb
9
awarded  Custodian
Feb
9
reviewed Satisfactory What should I be looking for in a heart rate monitor?
Dec
14
revised Continuous heart rate monitor
added link to review
Dec
14
suggested approved edit on Continuous heart rate monitor
Dec
1
comment Firstbeat Energy Expenditure Algorithm
I posted the question here because I am interested in an algorithm that I can run on my own HR data. The NN is nothing but a black box that gives more or less unverifiable and maybe unstable output after training with a huge dataset. As I don't have access to this training dataset, understanding the NN is not too useful for me. I am more interested in the physiological model + usable algorithm that depends only on a few coefficients. It is easier to understand the limitations of such a model, too.
Nov
22
comment Firstbeat Energy Expenditure Algorithm
I'd like to get a general understanding for the algorithms and their limitations. I am still reading Saalasti's thesis, it seems he is using a NN and trains it on the data, which gives a model that is not easy to understand and impossible to reproduce without the original dataset. For example all he states about EPOC is EPOC=f(EPOC_{t-1}, intensity, Dt) (107), what is really going on is hidden in the model. I am looking for algorithmic descriptions together with empirically obtained values to calculate things like Training Effect, EPOC, etc. myself.
Nov
22
revised Firstbeat Energy Expenditure Algorithm
added link to program
Nov
22
awarded  Editor
Nov
22
revised Firstbeat Energy Expenditure Algorithm
fixed links
Nov
22
comment Firstbeat Energy Expenditure Algorithm
Equation (106) gives relative VO2 as a function of the relative heartrate from a fit. pVO2 = 1.459*(pHR)*2 - 0.49*pHR+0.04. It's difficult to follow what the EPOC model really does with all the hidden parameters. I don't think they use a learning NN in a watch. From having a short look at the thesis, it looks like they use it as a black box function for a fit. Maybe someone with more knowledge about NNs can comment.
Nov
22
comment Firstbeat Energy Expenditure Algorithm
I found the thesis of one of the Firstbeat people jyx.jyu.fi/dspace/bitstream/handle/123456789/13267/…