Machine Learning
147 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
t-SNE
Nonlinear technique for visualizing high-dimensional data
AdaBoost
Combining weak classifiers with adaptive sample weighting
Ridge Regression
Regularized regression using L2 penalty to prevent overfitting
Markov Chain
Stochastic model where events depend only on previous state
Active Learning
Algorithm selecting most informative data points for labeling