Machine Learning
288 views
Gradient Descent
Quick Definition
Optimization algorithm minimizing loss by steepest descent
Full Definition
An optimization algorithm minimizing a loss function by moving in the direction of steepest descent.
Examples
neural network training, parameter optimization, loss minimization
Related Terms
backpropagation
learning-rate
optimization
More Machine Learning Terms
Loss Function
Math function measuring prediction vs actual value difference
A/B Testing
Statistical method comparing two versions for performance
Kernel Trick
Transforming data to higher dimensions for linear separability
Bias-Variance Tradeoff
The tension between model simplicity and complexity errors
Hyperparameter Tuning
Finding optimal configuration values before training
Confusion Matrix
Table visualizing classifier performance by class predictions