去中心化的室内定位众包数据质量控制方法

张学军,邝浚鑫,李成泽,李梅,张斌,加小红

Decentralized indoor positioning crowdsourcing data quality control method

Xuejun ZHANG,Junxin KUANG,Chengze LI,Mei LI,Bin ZHANG,Xiaohong JIA

表 8 PoW、PoS、DpoS、PoD与PoWFQ对模型性能的影响

Tab.8 Effect of PoW, PoS, DpoS, PoD and PoWFQ on model performance

共识算法 数据集 $ {a_{{{\mathrm{train}}}}} $ $ {l_{{{\mathrm{train}}}}} $ $ {a_{{{\mathrm{val}}}}} $ $ {l_{{{\mathrm{val}}}}} $ PoW UJIIndoorLoc 0.7710 0.7472 0.4448 3.7624 WiFi-RSS 0.7056 0.9821 0.6887 1.1992 MALL 0.8720 0.3458 0.9093 0.3323 PoS UJIIndoorLoc 0.8005 0.6661 0.6700 1.4324 WiFi-RSS 0.6999 0.9977 0.6962 1.2308 MALL 0.8832 0.3264 0.9036 0.3490 DPoS UJIIndoorLoc 0.7957 0.6799 0.6412 1.6682 WiFi-RSS 0.6779 1.0718 0.6544 1.3273 MALL 0.8716 0.3704 0.8957 0.3506 PoD UJIIndoorLoc 0.8432 0.5876 0.3526 5.6292 WiFi-RSS 0.3086 2.8100 0.0782 4.4087 MALL 0.8900 0.3425 0.2962 3.2239 PoWFQ UJIIndoorLoc 0.8919 0.3488 0.6563 2.1507 WiFi-RSS 0.7918 0.6556 0.5058 2.0796 MALL 0.9093 0.2617 0.8927 0.3614