---

Week 7: Journal Questions

Answer the following question for Monday's reading about convolution and deep learning (Mitchell Chapters 4-5)

1. Looking at the photo shown above, write several sentences about what you can infer about what is taking place in the photo.
2. Based on the reading, what might a convolutional, deep neural network be able to infer about this same photo?
3. On page 82, Mitchell writes that "the units in a ConvNet act as detectors for important visual features, each unit looking for a its preferred feature in a specific part of the visual field." How is this different from how a typical sigmoid unit operates in a standard layered neural network?

Answer the following question for Wednesday's reading about convolutional neural networks

1. What advantages do CNNs have over fully-connected networks? What disadvantages might they have?
2. What types of problems are CNNs well suited for? What types of problems might they be they poorly suited for?