13.07.2015 Views

Thesis - Instituto de Telecomunicações

Thesis - Instituto de Telecomunicações

Thesis - Instituto de Telecomunicações

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

128 CHAPTER 6. APPLICATIONS AND RESULTS6.2.4 Authentication ResultsThe EDA data is now used in the format of a sequential classifier in authentication mo<strong>de</strong>as <strong>de</strong>scribed in the previous chapter, in section 5.3.1.We also present the improvements resulting from fusing our EDA based classifier witha synthetic data classifier.Figure 6.11: EDA Equal error rate results for and increasing number of events.Figure 6.11 presents the results of the stand alone EDA biometric system with valuesstarting in 35% EER, for a 1 EDA event classifier, to near 10%, with 40 sequential EDAevents.Uncertainty Based Classification FusionThe low data separability led to an error probability that is unusable as a stand-alone biometricsystem. The two other signals, ECG and HCI, behave better if we have a sufficientlylarge sample sequence. We observe comparable results with respect to other behavioralbiometrics when longer sequences of data are available. In the case of the EDA signal,even with a sequence of samples, the error does not <strong>de</strong>crease to acceptable values. Theoption is data fusion with other, or sets of other, hard-biometric techniques. We need to

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!