Originally Posted by
wphamilton
Deep learning is primarily a sort of classification scheme, or high level abstractions from datasets if you prefer jargon. Training the classifieer to a good degree of accuracy, random forest or neural network or whatever, is more computationally intensive than a handlebar device can handle so I suspect that you're not really implementing real-time supervised learning. You're just running the heart rate et al and maybe some elements of the current workout through a classifier - right? - which is a cool idea but perhaps you shouldn't advertise it as AI with deep learning. If it's true that it's just a classifier.
That said, how much I'd want one depends on the details. What information does it specifically provide and how would that help me? Can I tell it to fire up some Joe Friel workout plan and have it tell me what's on the schedule for today? For any given moment in the ride? Can it evaluate my performance during the ride, in the context of the training plan, and modify the plan based on that?
Thanks for your reply on the post.
It does the following : The software engine takes into account three things
1.Geography
Factoring Terrain, weather and incline
2.Physiology
Factoring Fatigue, power and endurance
3.History
Factoring recent goals, recent performance and sleep
Depending on your goal criteria it will advice the cyclist. Current training systems miss the connection between how your performing vs what your actually training.
''For any given moment in the ride? Can it evaluate my performance during the ride, in the context of the training plan, and modify the plan based on that?''
Yes this is what i meant while stating real time training.
''What information does it specifically provide and how would that help me? Can I tell it to fire up some Joe Friel workout plan and have it tell me what's on the schedule for today? ''
It is hard for a startup company to get endorsement due to mainly financial constraints. I would like to know how much is endorsement by a popular athlete important to you?
Regards