Beginning Anomaly Detection Using

save_weights_only: If set to True, then only the weights will be 

saved. Essentially, if True, model.save_weights(filepath), else model.

save(filepath).

• 

mode: Can choose between auto, min, or max. If save_best_only is 

True, then you should pick a choice that would suit the monitored 

quantity best. If you chose val_acc for monitor, then you want to pick 

max for mode, and if you choose val_loss for monitor, pick min for 

mode.

• 

period: How many epochs there are between each checkpoint.

Now, you can train your model using code similar to Figure 

A-5

.

Appendix A   intro to KerAs

---

324

The model.fit() function has a big list of parameters:

• 

x: This is a Numpy array representing the training data. If you have 

multiple inputs, then this is a list of Numpy arrays that are all training 

data.

• 

y: This is a Numpy array that represents the target or label data. 

Again, if you have multiple outputs, then this is a list of target data 

Numpy arrays.

• 

batch_size: Set to 32 by default. This is the integer number of 

samples to run through the network before updating the gradients.

•