Explain like I'm five
Imagine you're baking cookies. The recipe tells you to set the oven temperature and baking time—those are your hyperparameters. You decide them before you put the cookies in, and they determine how well the cookies turn out.
Artificial Intelligence Difficulty 35/100
Hyperparameter
Turn the knob, bake the blob.
Hyperparameters are settings you choose before training an AI model that control how it learns, like the knobs on a stove that you set before cooking.
Why it matters
Hyperparameters matter because they directly affect how well an AI model learns and performs; picking the wrong ones can make the model useless. You encounter them whenever you train a neural network, like in image recognition or language models.
Common misconception
Many people think hyperparameters are learned by the model during training, like weights, but they are actually set manually beforehand. Another mistake is assuming one set of hyperparameters works for every problem, but they often need tuning for each specific task.
Formal definition
In machine learning, hyperparameters are configuration variables external to the model whose values are set before the learning process begins. They govern the training algorithm's behavior, such as learning rate, number of layers, or batch size, and are not updated during training. Choosing optimal hyperparameters often requires experimentation or automated search methods like grid search or Bayesian optimization.