Table 1 The table presents the average cross-validation accuracy results for different models, where ’MH Orig’ refers to the original MHEALTH data, ’MH Aug’ refers to the augmented MHEALTH data, and the same notation applies to PAMAP2

MH Orig

\(\begin{array}{l}91.70\\ (91.30)\end{array}\)

\(\begin{array}{l}88.15\\ (87.73)\end{array}\)

\(\begin{array}{l}93.36\\ (96.13)\end{array}\)

\(\begin{array}{l}90.49\\ (89.72)\end{array}\)

\(\begin{array}{l}58.11\\ (54.79)\end{array}\)

\(\begin{array}{l}94.57\\ (94.36)\end{array}\)

\(\begin{array}{l}81.06\\ (80.83)\end{array}\)

\(\begin{array}{l}87.21\\ (86.35)\end{array}\)

MH Aug

\(\begin{array}{l}86.72\\ (85.73)\end{array}\)

\(\begin{array}{l}92.04\\ (91.64)\end{array}\)

\(\begin{array}{l}93.51\\ (93.52)\end{array}\)

\(\begin{array}{l}94.98\\ (94.95)\end{array}\)

\(\begin{array}{l}84.19\\ (83.36)\end{array}\)

\(\begin{array}{l}91.92\\ (91.91)\end{array}\)

\(\begin{array}{l}90.34\\ (90.01)\end{array}\)

\(\begin{array}{l}91.77\\ (91.68)\end{array}\)

P2 Orig

\(\begin{array}{l}83.98\\ (84.06)\end{array}\)

\(\begin{array}{l}82.74\\ (82.57)\end{array}\)

\(\begin{array}{l}80.74\\ (80.89)\end{array}\)

\(\begin{array}{l}76.72\\ (76.47)\end{array}\)

\(\begin{array}{l}69.03\\ (67.38)\end{array}\)

\(\begin{array}{l}81.09\\ (81.32)\end{array}\)

\(\begin{array}{l}79.69\\ (79.57)\end{array}\)

\(\begin{array}{l}83.01\\ (82.96)\end{array}\)

P2 Aug

\(\begin{array}{l}80.74\\ (78.50)\end{array}\)

\(\begin{array}{l}86.61\\ (86.57)\end{array}\)

\(\begin{array}{l}82.21\\ (87.26)\end{array}\)

\(\begin{array}{l}89.91\\ (89.86)\end{array}\)

\(\begin{array}{l}80.82\\ (77.84)\end{array}\)

\(\begin{array}{l}88.10\\ (88.17)\end{array}\)

\(\begin{array}{l}85.39\\ (85.34)\end{array}\)

\(\begin{array}{l}87.34\\ (87.22)\end{array}\)