Python
This is the most concise code for incrementing the variable x.
What is x+=1?
In a classical search problem, this is the set of nodes next in line to be expanded.
What is the frontier?
Unlike classification tasks, tasks of this sort produce numbers from a continuous range as outputs.
What is regression?
This simple Boolean function cannot be computed by a single neuron because it is not linearly separable.
What is XOR?
An automatic soap dispenser at Facebook was accused of this -ism.
What is racism?
This is like a list, but immutable.
What is a tuple?
This preferred uninformed search algorithm combines the best features of depth-first and breadth-first search.
What is iterative deepening?
This sort of learning doesn't tell the learner the correct answer, but tells it how well it did.
What is reinforcement learning?
In Python,
1 / (1 + math.exp(-x)).
What is the logistic sigmoid function?
This term can refer either to discriminatory behavior displayed by an AI system or to the constant term in a linear function.
What is bias?
This is the value of the expression 'abcdef'[2:4].
What is 'cd'?
Given a consistent heuristic, this is the optimal search algorithm.
What is A* search?
This technique produces a function made of a hierarchy of if/else decisions.
What is a decision tree?
This technique, independently discovered by several researchers, allows multilayer neural networks to be trained.
What is backpropagation?
This term refers to a system that operates without human intervention.
What is autonomous?
This type of expression is often more concise than a for loop when converting a list into another list.
What is a list comprehension?
This enhancement to minimax search finds the same answer with less computation.
What is alpha-beta pruning?
When this happens, the learner is memorizing the noise in the training data.
What is overfitting?
This operation, applied in a small receptive field around each pixel of an image, greatly improved the effectiveness of neural networks for computer vision.
What is convolution?
This thought experiment argues that a system that behaves like a person might not understand what it is doing.
What is Searle's Chinese Room?
These methods, which often begin and end with double underscores, allow you to overload operators like + and ==.
What are magic methods?
This search technique, involving random sampling, was crucial in the development of AlphaGo.
What is Monte Carlo tree search?
This technique slices up the training data so as not to waste any when doing validation.
What is cross-validation?
This modification to gradient descent includes a fraction of the previous gradient.
What is momentum?
Natural language processing researchers were dismayed to learn that most of the Wikipedia entries in this language were actually written by an American teeenager who was doing an accent.
What is Scots?