US 12,254,421 B2
Machine learning knowledge management based on lifelong boosting in presence of less data
Anil Goyal, Dossenheim (DE); Ammar Shaker, Heidelberg (DE); and Francesco Alesiani, Heidelberg (DE)
Assigned to NEC CORPORATION, Tokyo (JP)
Filed by NEC Corporation, Tokyo (JP)
Filed on Dec. 6, 2023, as Appl. No. 18/530,331.
Application 18/530,331 is a continuation of application No. 16/889,825, filed on Jun. 2, 2020, granted, now 11,995,566.
Prior Publication US 2024/0127087 A1, Apr. 18, 2024
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06N 5/04 (2013.01) [G06N 20/00 (2019.01)] 13 Claims
OG exemplary drawing
 
1. A method for lifelong machine learning using boosting, the method comprising:
receiving a new task's learning sample and a knowledge base (KB);
learning weights over previously learned classifiers, and updating a distribution of weights over the learning sample;
using the updated distribution of the weights over the learning sample, sequentially learning a set of task-specific classifiers using a boosting algorithm and the distribution of the weights over the learning sample;
selecting less than 30% of samples from the learning sample which are hard to classify based on a corresponding distribution of weights over the learning sample and pruning some of the newly learned task-specific classifiers based on weights over the classifiers; and
updating KB based on new task, for old tasks, by learning the weights over the newly learned task-specific classifiers using old tasks hard examples.