Table 3 Summary of the predictive models.
From: Training load responses modelling and model generalisation in elite sports
Model | \(R^2\) | MAE | RMSE | Hyper parameters* |
|---|---|---|---|---|
\(\text {DR}_I^*\) | 0.206 ± 0.093 | 0.189 ± 0.055 | 0.225 ± 0.053 | \(\overline{k_1} = - 3.95e{-}05, k_1 \in [- 4.85e{-}05; - 3.19e{-}05]\) |
\(\overline{k_3} = - 7.75e{-}09, k_3 \in [- 4.01e{-}09; - 1.71e{-}08]\) | ||||
\(\overline{\tau _1} = 36.02, \tau _1 \in [25.82 ; 42.28]\), \(\overline{\tau _2} = 22.57, \tau _2 \in [14.58 ; 26]\), \(\overline{\tau _3} = 5.23, \tau _3 \in [4.33 ; 6.67]\) | ||||
\(\text {ENET}_I\) | 0.150 ± 0.010 | 0.169 ± 0.020 | 0.197 ± 0.023 | \({\overline{\alpha }} = 0.176, \alpha \in [0;0.6]\), \({\overline{\lambda }} = 0.273, \lambda \in [0;1]\) |
\(\text {PCR}_I\) | 0.164 ± 0.068 | 0.173 ± 0.025 | 0.201 ± 0.027 | \(\overline{n \, comp} = 1.918, n \, comp \in [1;3]\) |
\(\text {RF}_I\) | 0.193 ± 0.074 | 0.170 ± 0.023 | 0.199 ± 0.024 | \(\overline{m try} = 8.90, m try \in [1;17]\) |
\(\text {ENET}_G\) | 0.179 ± 0.063 | 0.150 ± 0.010 | 0.176 ± 0.012 | \(\alpha = 0.28, \lambda = 0.02\) |
\(\text {PCR}_G\) | 0.17 ± 0.053 | 0.22 ± 0.044 | 0.259 ± 0.062 | \(n comp = 3\) |
\(\text {RF}_G\) | 0.164 ± 0.069 | 0.163 ± 0.017 | 0.195 ± 0.017 | \(m try = 16\) |
