Hard#320
Fleet Recovery
Coordinate two robots with one shared policy under actuator stalls, limited batteries, and hidden layouts.
Author: Echipa științifică ONIAYour best score:N/A
Loading...
Available problems
Explore and solve machine learning challenges.
110 problems
Coordinate two robots with one shared policy under actuator stalls, limited batteries, and hidden layouts.
Build a cost-aware adaptive DAG for noisy ternary identification that remains robust under receiver shift.
Recognize known and unknown audio events, locate their boundaries, and remain calibrated under domain shift.
Design an adaptive decision graph that robustly identifies a transmitter in at most 15 calls.
Train a delivery policy that recovers from actuator stalls and generalizes to hidden layouts.
Detect overlapping sound events over time while adapting to new classes without forgetting the old ones.
Identify hidden animals in at most 15 calls by adaptively choosing yes/no questions and guesses.
Build an adaptive YES/NO question policy that identifies each animal from a catalogue of 250 candidates using at most 15 calls.
Train a behavioral-cloning model to deliver packages on unseen 8 × 8 maps.