**4. Results**

*Natural Hazards - Risk, Exposure, Response, and Resilience*

layer was implemented for this work.

popular and the first being the backpropagation, also known as generalized delta rule, popularized by [26]. The backpropagation learning algorithm is a supervised learning method and is an implementation of the Delta rule. It requires the desired output for any given input to be able to compute the output error. The main idea of the algorithm is to have a backward propagation of the errors from the output nodes to the inner nodes. For the construction of the backpropagation learning algorithm, we need to compute the gradient of the error of the network with respect to the network's modifiable weights. A DFANN network with 4 hidden layers and 12 neurons in each

**3.2 Recurrent neural networks with long short-term memory (RNN-LSTM)**

In addition, they classify and predict based on time series data, since there may be delays of unknown duration between important events in a series of time. It allows clearly remembering events selected from far away in the past, which contrasts with basic NRs, for which the memory of an event decays over time [27].

As firstly proposed by Rumelhart [26], recurrent neural networks have a primitive type of memory, in the form of recurrent layers that can operate in time [27]. Each recurrent layer takes both the output of the previous layer and an internal output of the current layer as inputs. Thus, RNNs are ideal for dealing with time series data [27]. RNNs can solve the purpose of sequence handling to a great extent but not entirely; they are great when it comes to short contexts, but to be able to build a story and remember it, the models need to be able to understand and remember the context behind longer sequences, just like a human brain. This is not possible with a simple RNN. Long short-term memory (LSTM) networks [28] are a type of RNN precisely designed to escape the long-term dependency issue of recurrent networks. LSTM recurrent networks (RNN-LSTM) have memory cells that have an internal recurrence (a self-loop), in addition to the outer recurrence of the RNN. The latter adds a nonlinear transformation to the inputs [28]. These memory cells, A, are controlled mainly by the memory door, the forgetting door (*ht*), and the output door. The memory door activates the entry of information to the memory cell, and the forgetting door selectively erases certain information in the memory cell and activates the storage to the next entry [29]. Finally, the output door decides what information the memory cell will emit [30]. The LSTM network structure is illustrated in **Figure 4**. Each cell has three gate activation functions σ and two output activation functions defined by tanh as a nonlinear

**8**

**Figure 4.**

*LSTM cells structure, based on the work by [31].*

transfer function.

**Figure 5** shows GIF estimation for the data preprocessing module, estimated for magnitudes >3, >4, >5, and >6, respectively. Note that with higher magnitudes, the GIF time series become thinner, due to the decrease of seismic events that fit in the category.

The structure implemented for both DFANN and RNN-LSTM models is shown in **Figure 6**.

The DFANN model performs slightly better than the RNN-LSTM models, in particular for lesser magnitudes (>3). **Table 1** shows the training and test performance measures (root mean square error, RMSE) for each magnitude group and DL model. Both models show better performances with magnitude >3, that is, when more information are available.

Also, a representation of the training and test results for the best model are shown in **Figure 7**. The model captures the trend very well; however, it does not perform accordingly in terms of the magnitude of the intensity function.

**Figure 5.** *Ground intensity function (GIF) estimation.*

#### **Figure 6.**

*Structure for the DL models, for both DFANN (on the left) and RNN-LSTM (on the right).*

**Figure 7.** *Training and test groups for the best model (DFANN, Mag > 3).*


#### **Table 1.**

*Root mean square error (RMSE) of the training and test groups for each DFANN and RNN-LSTM deep learning models.*

**11**

this work.

*Assessing Seismic Hazard in Chile Using Deep Neural Networks*

seismic areas such as the local ETAS models [7, 11].

variables and their uncertainty.

**6. Conclusion**

prediction approaches.

**Acknowledgements**

and allowing to "generate" new prediction seismic risk maps.

This work introduces a novel approach to predict the temporal ETAS-GIF alternative to the statistical approach proposed by [14]. The deep learning method has recently been used for predicting locations of aftershock events [31] especially based on ground motion data. The first use of a feedforward neural network for the

Possible extensions of the deep learning approach could be to include the ground motion together to other variables [30, 31] as inputs of the model and to incorporate the spatial dimension for a spatiotemporal prediction [33–35]. Some statistical techniques could be used for identifying possible patterns and inputs [36–37]. Also, since seismic events could be characterized by different features depend-

prediction of seismic hazard was introduced by [32] in the spatial domain.

ing of the different locations of the principal events, we think that DL neural network models could be used for characterizing earthquakes in some specific

Different neural networks models could be used for comparing earthquake predictions [38]. For example, Bayesian DL neural networks could be used for a new prediction scenario considering the uncertainty of major earthquake occurrences and the probability of recurrence in a similar way to the Bayesian approach proposed by [32]. Additionally, other DL and machine learning approaches as convolutional neural networks (CNN), generative networks (GN), and random forest regression (RFR) could be implemented by incorporating the spatial component

However, the main limitation of neural networks is that they are considered "black boxes" since it is difficult to quantify the correlation between the involved

This chapter deals with the estimation of seismic risk given by the temporal ETAS conditional intensity function. To achieve this goal, two deep learning models were implemented: a deep feedforward artificial neural network and a recurrent long short-term memory network. The results show a good estimation, in particular with the DFANN model. However, it should be pointed out that both implemented models could be improved by adding more hidden layers or stacking more LSTM layers in the DFANN and RNN-LSTM models, respectively. Also, exogenous variables (such as ground motion among others) could be considered for improving the predictions. Since the proposed model only considers a temporal model, extensions to the prediction of earthquake locations will be considered in future works. We think that deep learning algorithms could be useful tools for many earthquake

The authors thank the National Research Center for Integrated National Disaster Management (CIGIDEN), CONICYT/FONDAP/15110017 (Chile) and CONICYT PFCHA/DOCTORADO BECAS CHILE/2018 – 21182037 for financing

*DOI: http://dx.doi.org/10.5772/intechopen.83403*

**5. Discussion**
