Table 1 Summary of results of models predicting the vehicle travel time after 6 h on our study’s street.

From: Predicting vehicle travel time on city streets for trip preplanning and predicting heavy traffic for proactive control of street congestion

 

AdaBoost 

Neural net. 

Gradient boos. 

Nearest neighbors 

Support vect. mach.

Mean absolute error

0.0046

0.0612

0.0337

0.0007

0.1817

R squared score

0.9993

0.9994

0.9996

0.9998

0.9898

Explained variance score

0.9993

0.9994

0.9996

0.9998

0.9898

Mean squared error

0.0115

0.0101

0.0056

0.0030

0.1770

Median absolute error

0.0000

0.0610

0.0222

0.0000

0.1000

Mean abs. percent. error

0.0005

0.0067

0.0035

0.0000

0.0166

Max error

3.8290

3.9016

3.8984

4.0000

4.7504

First important predictor

t0

t18

t0

t1

t0 & t6

Second important predictor

t14 & t18

t0 & t1

t18

t0

t2, t7, & t9