Methods and apparatus for coaching based on workout history
System and method for coaching based on workout history. Improved solutions enable intelligent management of a user's personal fitness journey based on workout recommendations that closely align with the user's traits. In one exemplary embodiment, workout data for a population of different individuals is analyzed to identify groups of similarly performing individuals. Each group of individuals is analyzed to generate an expected profile that approximates the physiological and/or psychological traits of the group. An expected profile includes heuristics and/or performance metrics that enable dynamic coaching during workouts. Subsequently thereafter, user's can be dynamically coached by their client device, based on the expected profile.
1. A method for enabling dynamic coaching feedback at a client device, comprising:
generating a plurality of expected profiles for a plurality of data records within a user workout history database by:
using a supervised machine learning model to group users into the plurality of expected profiles; and
using unsupervised machine learning to generate heuristics and performance metrics for each of the plurality of expected profiles;
associating a first user with one of the expected profiles;
receiving user input from the first user relating to at least one workout;
in response to receiving the user input, recommending a workout for the first user based on the expected profile associated with the first user;
wherein the expected profile associated with the first user includes at least one heuristic for generating dynamic feedback with the workout, wherein the at least one heuristic for generating the dynamic feedback comprises a rule for modifying a second exercise of the workout based on an actual performance of a first exercise of the workout, wherein the expected profile comprises an expected performance level for the first exercise of the workout, and wherein said expected performance level is based on a model profile taken from a subset of users identified as having a similar physiology and similar psychology as the first user; and
updating a user data record of the first user based on a logged performance corresponding to the at least one workout.
2. The method of claim 1 , where the workout data records are obtained from a population of users.
3. The method of claim 2 , wherein the population of users are categorized based on at least one physiological or psychological trait.
4. The method of claim 1 , where the at least one heuristic for generating the dynamic feedback comprises a rule for motivating the user based on an actual performance of the first exercise of the workout.
5. The method of claim 1 , further comprising identifying the expected profile from the plurality of expected profiles based at least in part on an assessment performance, the method further comprising disassociating the first use with the expected profile and associating the user with a different one of the plurality of expected profiles based at least in part on the logged performance.
6. The method of claim 1 wherein the similar psychology is one or more of a similar confidence state, motivational state, or emotional state.
7. A health tracking server configured to recommend workout to users, comprising:
a network interface;
a processor;
a non-transitory computer-readable medium comprising one or more instructions, which when executed by the processor, causes the health tracking server to:
obtain workout data records associated with users;
analyze the workout data records to group the users into a plurality of distinct subsets, wherein the users are grouped into the subsets based on machine learning data analysis of physiological and psychological traits;
analyze each subset to generate a corresponding model profile for each subset, each model profile comprising at least a heuristic and a performance metric associated therewith, wherein the at least one heuristic is a heuristic for generating dynamic feedback comprising a rule for modifying a second exercise of a workout based on an actual performance of a first exercise of the workout;
receive user input from a user device relating to at least one first exercise of a workout of a first user;
match the first user with a first model profile based at least in part on said received user input relating to at least one first exercise;
recommend to the first user a modification to at least one second exercise of the workout of the first user based on the first model profile, wherein the modification is recommended responsive to said user input;
receive subsequent user input from the user device relating to the at least one second exercise of the workout of the first user; and
based at least in part on the received user input and subsequent user input, use machine learning to identify a pattern of the first user that identifies the first user with a second model profile; and
match the first user with the second model profile.
8. The health tracking server of claim 7 , wherein the corresponding model profile for each subset is generated based on machine learning data analysis of physiological and psychological traits.
9. The health tracking server of claim 7 , wherein the first user is matched with the first profile from a plurality of profiles based at least in part on an assessment exercise.
10. The health tracking server of claim 7 , wherein the first user is matched with the first profile from a plurality of profiles based at least in part on a history of user workout data records.
11. The health tracking server of claim 7 , where in the first user is matched with the first profile from a plurality of profiles based at least in part on a fitness goal identified by the first user.
12. The health tracking server of claim 7 the psychological traits include one or more of an assessed confidence, motivation or emotional state.
13. A user apparatus, comprising:
a user interface configured to receive user input related to a workout performed by a user;
a network interface configured to transmit the user input to a health tracking server;
a processor; and
a non-transitory computer-readable medium comprising one or more instructions, which when executed by the processor, causes the user apparatus to:
receive a first recommended workout and a first model profile comprising a performance metric for a first exercise of a workout, wherein the performance metric includes least one heuristic for generating dynamic feedback for modifying a second exercise of the workout based on an actual performance of the first exercise of the workout, and wherein said recommended workout is received from a health tracking server configured to:
(i) match the user with the first model profile based at least in part on said user input related to the workout performed by the user; and
(ii) generate said recommended workout responsive to said user input;
(iii) use machine learning to identify a pattern of the user that identifies the user with a second model profile; and
(iv) match the user with the second model profile;
monitor performance during the recommended workout;
when the performance does not match the performance metric for the first exercise, provide dynamic feedback for the second exercise based on the first user profile; and
receive a second recommended workout and the second model profile.
14. The user apparatus of claim 13 , wherein the recommended workout is selected from a plurality of recommended workouts.
15. The user apparatus of claim 13 , wherein the first user profile and second user profile are matched for the user associated with the user apparatus.
16. The user apparatus of claim 13 , wherein the dynamic feedback comprises a revised workout.
17. The user apparatus of claim 16 , wherein the one or more instructions, when executed by the processor, causes the user apparatus to log a revised performance with the revised workout.
18. The user apparatus of claim 13 , wherein the dynamic feedback comprises a motivational message.