Translation guide
In machine learning and statistics, overfitting occurs when a model learns the training data too well, capturing noise rather than the underlying pattern, and thus performs poorly on new data. This guide explains how to express this concept in Japanese.
The standard technical term for overfitting in data science contexts.
The most common and standard translation for 'overfitting' in machine learning. Literally 'excessive learning'.
This model is overfitting.
Also used, especially in statistical contexts. Literally 'excessive fitting'.
The loanword from English, commonly used in technical discussions.
We need to solve the problem of overfitting.
In Japanese technical writing, 過学習 is the most standard term. オーバーフィッティング is also widely understood, especially in spoken or informal technical contexts. 過適合 is less common but appears in some statistical literature.
We apply regularization to prevent overfitting.
Overfitting reduces the model's generalization performance.