This glossary defines general machine learning terms as well as terms specific to TensorFlow. Below is a small selection of the most popular entries. You can access this glossary here. For other related glossaries, follow this link.

- A/B testing
- activation function
- agglomerative clustering
- artificial general intelligence
- AUC (Area under the ROC Curve)
- backpropagation
- bag of words
- Bayesian neural network
- Bellman equation

- binning
- boosting
- centroid-based clustering
- class-imbalanced dataset
- clustering
- collaborative filtering
- confusion matrix
- convenience sampling
- convex optimization
- convolutional filter
- convolutional neural network
- cross-entropy
- cross-validation
- data augmentation
- Dataset API (tf.data)
- decision boundary
- decision tree
- deep neural network
- Deep Q-Network (DQN)
- dimension reduction
- discriminative model
- ensemble
- false positive (FP)
- feature engineering
- feature extraction
- federated learning
- feedforward neural network (FFN)
- generalization curve

- generalized linear model
- generative adversarial network (GAN)
- gradient descent
- hashing
- hidden layer
- hierarchical clustering
- hinge loss
- hyperparameter
- input layer
- Keras
- Kernel Support Vector Machines (KSVMs)
- k-means

- k-median
- L1 loss
- L1 regularization
- learning rate
- least squares regression
- linear model
- linear regression
- logistic regression
- loss curve
- Markov decision process (MDP)
- Markov property
- matrix factorization
- Mean Squared Error (MSE)
- minimax loss
- model training
- multinomial classification
- natural language understanding
- neural network
- N-gram
- noise
- normalization
- NumPy
- objective function
- outliers
- overfitting
- pandas
- perceptron

- prediction bias
- pre-trained model
- prior belief
- proxy labels
- Q-function
- quantile
- quantile bucketing
- random forest
- recommendation system
- recurrent neural network

- regression model
- regularization
- reinforcement learning (RL)
- ridge regularization
- sampling bias
- scaling
- scikit-learn
- scoring
- selection bias
- semi-supervised learning
- sentiment analysis
- sigmoid function
- similarity measure
- size invariance
- sparse feature
- stationarity
- stochastic gradient descent (SGD)
- structural risk minimization (SRM)
- subsampling
- supervised machine learning
- synthetic feature
- TensorFlow
- time series analysis
- training set
- transfer learning
- true positive (TP)
- underfitting
- unsupervised machine learning
- validation
- vanishing gradient problem
- Wasserstein loss
- Weighted Alternating Least Squares (WALS)