Chapter 11 – Combining Reinforcement Learning and Deep Learning
- A CNN can be trained to understand an abstract concept? (Yes | No)
Yes. A CNN can classify images and make predictions. But CNNs can analyze any type of object or representation. An image, for example, can be linked to a word or phrase. The image thus becomes a message in itself.
- Is it better to avoid concepts and only use real-life images? (Yes | No)
No. Images provide many practical applications, but at some point, more is required to solve planning problems, for example.
Planning requires much more than this type of dataset.
- Do planning and scheduling mean the same thing? (Yes | No)
No. Planning describes the tasks that must be carried out. Scheduling adds a time factor. Planning tells us what to do, and scheduling tells us when.
- Is Amazon's manufacturing patent a revolution? (Yes | No)
No. Manufacturing clothing...
