When it comes to machine learning, it is algorithms, not computer programmers, that create the rules that a computer must follow. Algorithms (A) [resemble / are resembling] the kinds of rules that parents set for their kids. They give computers instructions that allow them to learn from data and, as a result, the computers can be taught to perform new, complex tasks without any additional manual programming. Machine learning basically works in the following way: training data is fed into a learning algorithm, (B) [it / which] then uses this input to create a new set of rules. In a sense, this is a way of generating an entirely new algorithm that can be referred to as the machine learning model. Through the use of different training data, the same learning algorithm could generate different models, from ones that teach the computer to (C) [translate / translating] languages to others that predict the behavior of the stock market.