Home › Per Daniel Harabor's follow-up work and the publ…
Per Daniel Harabor's follow-up work and the public literature, JPS+ (the offline variant) adds which optimisation over base JPS?
APreprocessed cardinal jump distances per cell
BRandom precomputed paths for fallback only
CA neural-network policy for direction picking
DStochastic A* with simulated annealing search
Answer & Solution
Correct answer: A. Preprocessed cardinal jump distances per cell
Harabor + Grastien 2014 paper + GameAIPro chapter: JPS+ stores 'jump' tables per cell at load time, removing online jump cost. 10–100× faster vs JPS.
Related questions
In Reynolds-style steering behaviours combined into a STEERING PIPELINE, the OUTPUT accelePer Dave Mark + Brian Schwab, a RESPONSE CURVE in utility AI maps an input variable to a sIn modern game AI, REINFORCEMENT LEARNING agents (e.g. OpenAI Five, DeepMind AlphaStar) lePer AlphaGo (Silver et al. 2016, Nature), MONTE CARLO TREE SEARCH (MCTS) explores a game tPer Daniel Harabor (2011 AAAI paper), JPS (Jump Point Search) accelerates A* on UNIFORM-COIn tactical AI, an INFLUENCE MAP is best described as which data structure?In behaviour trees beyond sequence + selector, a DECORATOR node typically does which?Per Dave Mark ('Behavioral Mathematics for Game AI'), UTILITY AI selects actions by which