Information Technology Reference
In-Depth Information
Algorithm 3.
CLPs
Accuracy Promotion
Input:
CLPs
,RFIDdataset
Output:
CLPs
with modified ones
1:
for
each
clp ∈ CLPs
do
2:
if
clp
has RFID record
then
if
clp
doesn't equal the
LP
of RFID information
then
3:
4:
clp ₐ LP
of RFID information;
5:
Mark
clp
as modified;
6:
end if
7:
end if
8:
end for
9:
return
CLPs
;
is a control group in experiments. Five metrics are applied to evaluate the ve-
racity: sensitivity, specificity, precision, accuracy and ROC (receiver operating
characteristic) curve.
4.1 Experiment Data Sets and Evaluation Metrics
To evaluate the veracity and time consuming of HelpMe, real datasets are adopted
in the experiments. Twenty representative SBSs in a primary road are selected
to be experimental data source. And the data sets cover data collected during
both day and night with different weather conditions. Data sets contain about
51,794,728 records of ALPR data and 33,971,265 records of RFID data ranging
from Feb. 1st to Mar. 1st in Nanjing.
Sensitivity is referred to as the true positive (recognition) rate, while speci-
ficity is the true negative rate. In addition, precision is used to assess the per-
centage of tuples labeled as real that actually are real tuples. And accuracy is
the combination of sensitivity and specificity. ROC curve is a useful visual tool
for comparing binary classification model. A ROC curve shows the trade-off be-
tween sensitivity and specificity for a given model. The area under the ROC
curve is a measure of the accuracy of the model.
4.2 Experiment Context
Technically, our experiments are conducted in a HANA cluster environment.
The overview of the experiment context is depicted in Fig. 5, and the services
recruited in our experiment are distributed among the cluster. And specific hard-
ware and software configurations are as following:
-
Forhardware,theclientisLenovoThink-padT430machinewithInteli5-3210M
2.50GHz processor, 4GB RAM and 250GB Hard Disk. And for master(one
node) of HANA cluster is HP Z800 Workstation Intel(R) Multi-Core X5690
Xeon(R),3.47GHz/12MCache,6cores,2CPUs,128GBDDR31066MHzECC
Reg RAM, 2TB 7.2K RPM SATA Hard Drive, for slave(one node) of HANA
cluster is HP Z800 Workstation Intel(R) Multi-Core X5690 Xeon(R),
