Lưu ý: Bản dịch của mục này hiện đang được kiểm tra chất lượng, vì vậy một số nội dung tạm thời chỉ hiển thị bằng tiếng Anh.
Mục từ này chưa được dịch sang ngôn ngữ của bạn, vì vậy nội dung gốc được hiển thị bên dưới.
underfitting
This term is a technical descriptor used almost exclusively within the fields of data science, statistics, and machine learning. It describes a failure of a model to learn the training data, typically because the model is too rigid or simple. This is contrasted with overfitting, where a model learns the noise in the data rather than the signal.
When used as a noun, it refers to the state or phenomenon of the error. When used as an adjective, it describes the specific model or the nature of the error. Because it is a specialized technical term, it is rarely used in general conversation outside of a professional or academic context involving predictive modeling.
Ý nghĩa
Ví dụ
The accuracy is terrible because we are dealing with severe underfitting.
I think the model is underfitting the training data.
Avoid underfitting by increasing the number of layers in your neural network.
Is this a case of overfitting or underfitting?
Is this a case of overfitting or underfitting?
The simple linear regression resulted in underfitting since the relationship was non-linear.
The simple linear regression resulted in underfitting since the relationship was non-linear.
We must reduce the regularization strength to stop the underfitting.
The model is too basic and is clearly underfitting.
Maybe I should add more features to prevent underfitting.