US 11,991,531 B2
Communication load forecasting accuracy with adaptive feature boosting
Chengming Hu, Montreal (CA); Xi Chen, Montreal (CA); Ju Wang, Montreal (CA); Hang Li, Montreal (CA); Jikun Kang, Montreal (CA); Yi Tian Xu, Montreal (CA); Xue Liu, Montreal (CA); Di Wu, Montreal (CA); Seowoo Jang, Seoul (KR); Intaik Park, Seoul (KR); and Gregory Lewis Dudek, Westmount (CA)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed on Jan. 7, 2022, as Appl. No. 17/570,767.
Claims priority of provisional application 63/174,872, filed on Apr. 14, 2021.
Prior Publication US 2022/0338019 A1, Oct. 20, 2022
Int. Cl. H04W 16/22 (2009.01); G06N 3/047 (2023.01); H04L 41/14 (2022.01); H04L 41/147 (2022.01); H04L 41/16 (2022.01); H04W 24/02 (2009.01); H04W 24/08 (2009.01)
CPC H04W 16/22 (2013.01) [G06N 3/047 (2023.01); H04L 41/145 (2013.01); H04L 41/147 (2013.01); H04L 41/16 (2013.01); H04W 24/02 (2013.01); H04W 24/08 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving a first dimension set;
extracting a first latent feature set from the first dimension set;
training a first base predictor based on the first latent feature set;
generating a second dimension set based on the first dimension set, the second dimension set having fewer dimensions than the first dimension set;
extracting a second latent feature set from the second dimension set;
training a second base predictor based on the second latent feature set; and
generating a traffic prediction based on the first base predictor and the second base predictor,
wherein generating the second dimension set comprises:
sampling each dimension in the first dimension set with a predetermined number of runs; and
determining that dimensions in the first dimension set with a number of occurrences in the predetermined number of runs that is greater than a sampling threshold are to be included in the second dimension set.