Recognizing seismic waves immediately is very important for the realization<br />
of efficient disaster prevention. Generally these systems consist of a network<br />
of seismic detectors that send real time data to a central server. The server<br />
elaborates the data and attempts to recognize the first signs of an earthquake.<br />
The current problem with this approach is that it is subject to false alarms. A<br />
critical trade-off exists between sensitivity of the system and error rate. To<br />
overcame this problems, an artificial neural network based intelligent learning<br />
systems can be used. However, conventional supervised ANN systems are difficult<br />
to train, CPU intensive and prone to false alarms. To surpass these problems,<br />
here we attempt to use a next-generation unsupervised cortical algorithm HTM.<br />
This novel approach does not learn particular waveforms, but adapts to<br />
continuously fed data reaching the ability to discriminate between normality<br />
(seismic sensor background noise in no-earthquake conditions) and anomaly<br />
(sensor response to a jitter or an earthquake). Main goal of this study is test<br />
the ability of the HTM algorithm to be used to signal earthquakes automatically<br />
in a feasible disaster prevention system. We describe the methodology used and<br />
give the first qualitative assessments of the recognition ability of the<br />
system. Our preliminary results show that the cortical algorithm used is very<br />
robust to noise and that can successfully recognize synthetic earthquake-like<br />
signals efficiently and reliably.