MY ONE BLOGby Manishdai
Artificial Intelligence

What is Overfitting?

Quick answer

When a machine learning model learns its training data too closely, including noise, and performs poorly on new, unseen data.

Overfitting, explained

An overfitted model is like a student who memorizes past exam answers instead of understanding the subject: it scores perfectly on familiar questions but struggles with new ones. In practice, it shows very high accuracy on training data and noticeably lower accuracy on validation or test data.

Overfitting is more likely when a model is very complex relative to the amount of training data. Common remedies include gathering more data, simplifying the model, regularization techniques that penalize complexity, dropout in neural networks, early stopping during training, and cross-validation. The opposite problem, underfitting, happens when a model is too simple to capture real patterns. Good machine learning practice balances the two.

Related terms

Hyperparameter

A setting chosen before training a machine learning model, such as learning rate or number of layers, that shapes how the model learns.

Machine Learning

A branch of AI in which computers learn patterns from data to make predictions or decisions without being explicitly programmed for each case.

Model Evaluation

The process of testing an AI model against defined tasks and criteria to measure how accurate, reliable, and safe it is.

Regression

A statistical and machine learning method for predicting a continuous numerical value based on the relationship between variables.

Guides that mention Overfitting