3.2.7 Benchmark with a "naive" classifier

We remark the following for the "naive" classifier (Table 19):



#### Table 16.

Best parameters maximizing precision+recall rate for different futures and first bar price structure allowing lower bounds for θMIR and higher bounds for θVPIN.


#### Table 17.

Best parameters maximizing precision+recall rate for different futures and median bar price structure allowing lower bounds for θMIR and higher bounds for θVPIN.


#### Table 18.

Best parameters maximizing precision+recall rate for different futures and mean bar price structure allowing lower bounds for θMIR and higher bounds for θVPIN.

An Assessment of the Prediction Quality of VPIN DOI: http://dx.doi.org/10.5772/intechopen.86532


Table 19.

• There are changes only for NQ and YM instruments in, respectively, last, median, and mean bar price structures and first bar price structure, where

• On average median bar price structure has the best precision+recall rate.

Futures Recall Precision Precision+recall θMIR n ω (buckets) Classifier θVPIN ES 0.9404 0.9402 1.8806 0.015 30 2500 Gaussian 0.99 EC 0.9127 0.9681 1.8808 0.006 30 2500 Student 0.99 CL 0.9232 0.9728 1.8960 0.016 30 2500 Student 0.99 NQ 0.8291 0.9833 1.8124 0.01 30 2500 Student 0.99 YM 0.9341 0.9872 1.9213 0.015 50 2500 Gaussian 0.999

Best parameters maximizing precision+recall rate for different futures and first bar price structure allowing

Futures Recall Precision Precision+recall θMIR n ω (buckets) Classifier θVPIN ES 0.9499 0.9498 1.8997 0.015 30 2500 Student 0.99 EC 0.9037 0.9717 1.8754 0.006 30 2500 Student 0.99 CL 0.9265 0.9718 1.8983 0.016 30 2500 Student 0.99 NQ 0.8881 0.9525 1.8406 0.02 30 2500 Student 0.999 YM 0.9829 0.9427 1.9256 0.015 30 2500 Gaussian 0.99

Best parameters maximizing precision+recall rate for different futures and median bar price structure allowing

Futures Recall Precision Precision+recall θMIR n ω (buckets) Classifier θVPIN ES 0.9526 0.9454 1.8980 0.015 30 2500 Student 0.99 EC 0.9058 0.9691 1.8749 0.006 30 2500 Student 0.99 CL 0.9302 0.9670 1.8972 0.016 30 2500 Gaussian 0.99 NQ 0.9188 0.9150 1.8338 0.02 40 2500 Gaussian 0.999 YM 0.9446 0.9779 1.9225 0.015 60 2500 Gaussian 0.99

Best parameters maximizing precision+recall rate for different futures and mean bar price structure allowing

We remark the following for the "naive" classifier (Table 19):

• It has worse results than VPIN on ES and YM cases.

• It has comparable results than VPIN on NQ case.

• There is no general trend for precision or recall rates with the increase of θVPIN.

θVPIN equals 0.999.

Table 16.

Table 17.

Table 18.

68

3.2.7 Benchmark with a "naive" classifier

Advanced Analytics and Artificial Intelligence Applications

lower bounds for θMIR and higher bounds for θVPIN.

lower bounds for θMIR and higher bounds for θVPIN.

lower bounds for θMIR and higher bounds for θVPIN.

Best parameters maximising precision+recall rate for different futures for the naive classifier allowing lower bounds for θMIR.


We may partially conclude that:

