Artificial Intelligence Revision

What's the best answer: similar to the human brain, answers questions without explicit programming, uses lots amounts of training data, often uses artificial neural networks.

• artificial intelligence

• deep learning

• machine learning

• expert systems

What's the best answer: intelligent machine that thinks and behaves similar to a human

• artificial intelligence

• deep learning

• machine learning

• expert systems

What's the best answer: system learns without directly having to be programmed

• artificial intelligence

• deep learning

• machine learning

• expert systems

What is this thing called?

The accuracy of neural networks are often improved by adding more ______ layers.

If you wanted to make a neural network that detects cats vs dogs vs rabbits from images. How many output nodes should you have?

Neural networks often have problems of overfitting that come from too many ______ layers

Another word for overfitting would be....

• Too much complexity to match input data

• The result is too tight and doesn't allow flexibility

• The result is too loose and accepts data points that are inaccurate.

• Not enough input data to train the model

Answer explanation

over-generalization. Overfitting occurs when a model is too complex and captures too much of the noise in the data, resulting in a model that performs well on the training data but poorly on unseen data. Over-generalization is the result of a model that is too simple and does not capture enough of the underlying patterns in the data, resulting in a model that performs poorly on both the training and unseen data.

One disadvantage of machine learning is that

• It requires lots of input data

• The results are less accurate than explicitly programmed models

• It is difficult to interpret the results

• The results don't work well in the real world

Another disadvantage of machine learning is that .....

• Training the model takes time

• You need very expensive computers to train the data

• Programmers have to write the statistical model by scratch

• It is expensive to implement