Mean Absolute Error (MAE)

MAE = 0

model is perfect

MAE = min

model is optimal

0 < MAE

model is less perfect

Relative Root Mean Square Error (RRMSE)

RRMSE = 0

model is perfect

RRMSE = min

model is optimal

Model Efficiency (EF)

EF = 1

model is perfect

EF = max

model is optima

EF < 1

model is less perfect

EF = −¥

model has no prediction capability

Coefficient of Residual Mass (CRM)

CRM = 1

model has no prediction capability

CRM < 1

model has some at least prediction capability

CRM close to 0

model is optimal

Coefficient of Determination (CD)

CD = 0

model has no prediction capability

0 < CD

model has some at least prediction capability

CD = max

model is optimal

Goodness of Fit (R²)

R² = 1

model is perfect

R² = max

model is optimal

R² = 0

model has no prediction capability