Machine Learning
284 views
Overfitting
Quick Definition
When a model learns training data noise and fails on new data
Full Definition
A modeling error where a model learns noise in training data, negatively impacting performance on new data.
Examples
high variance, poor generalization, complex models
Related Terms
regularization
cross-validation
underfitting
More Machine Learning Terms
Backpropagation
Algorithm computing gradients for neural network weight updates
Dropout
Regularization randomly deactivating neurons during training
UMAP
Fast dimensionality reduction preserving local and global structure
Decision Tree
Tree-like model splitting data based on feature values
Semi-Supervised Learning
Combining small labeled datasets with large unlabeled data
Hidden Markov Model
Statistical model with hidden states and observable outputs