Learning and Mining using Bagged Augmented Decision Trees (LAMBAST)

Standoff weapons like seeker missiles enable strikes at targets of opportunity in limited-access areas while minimizing risk to warfighters. However, seekers cannot be deployed quickly enough for these short-notice missions because of two limitations of their onboard ATR systems: 1) ATRs cannot learn from novel, mission-specific data after initial training is complete; and 2) ATRs cannot identify pertinent, non-redundant information from large training databases. These limitations make it impossible to train a seeker’s ATR in a feasible timeframe for short-notice missions. To remedy these problems and make short-notice seeker missions a reality, we propose LAMBAST, or Learning and Mining using Bagged Augmented Decision Trees. LAMBAST examines large databases and extracts sparse, representative subsets of target and clutter samples of interest. For data mining, LAMBAST uses a variant of decision trees, called random decision trees (RDTs). RDTs are immune to overfitting and can incorporate novel, mission-specific data after initial training via perpetual learning. We augment these trees with a distribution modeling component that eliminates redundant information, ignores misrepresentative class distributions in the database, and stops training when decision boundaries are sufficiently sampled. These augmented random decision trees enable fast construction of reliable, mission-specific ATR and make short-notice seeker missions possible.