7 ms·
You can never use 1950s control theory to solve your problem? I think you didn't understand the my comment, so please let me clarify: I was claiming that even t
by gugagore 10y ago
You can never use 1950s control theory to solve your problem? I think you didn't understand the my comment, so please let me clarify: I was claiming that even the control problems in this RC-car-lane-keeping domain can benefit from learning approaches.