30.9.2. Deep Learning Architectures

Try these first

1. 1.

   What does CNN stand for?

   Hint

   Think about networks used for images.

2. 2.

   Which architecture is designed for sequence data?

   Hint

   Recall the type that considers the prior order of data.

3. 3.

   What are CNNs primarily used for?

   • Image processing
   • Sound analysis
   • Text processing

   Hint

   Think of visuals and structural analysis.

4. 4.

   True or False: LSTM networks are a type of CNN.

   • True
   • False

   Hint

   Recall the specific roles of these networks.

5. 5.

   Design a simulation where a CNN analyzes images of structural defects. What data would you need, and how would you measure accuracy?

   Hint

   Consider the importance of accurate labeling in training data.

6. 6.

   Compose an argument for using LSTM networks over traditional predictive methods in monitoring structural integrity over time. What advantages do they offer?

   Hint

   Think of a time series as a flowing river, where past states can influence future ones.