ML
Computers teaching themselves
ML is the abbreviation for machine learning, the branch of artificial intelligence where computers improve at tasks through experience rather than explicit programming: The app uses ML to recommend songs you'll like.
In other contexts, ML can mean milliliter or markup language (as in HTML).
Layered computational learning systems
- deep learningLayered neural network methods
- reinforcement learningTraining through reward signals
- classificationSystematic grouping by traits
- regressionReturn to former state
clustering · decision trees · backpropagation · activation function · k-means
Computational analysis of speech
- NLPMachine language comprehension
- computer visionAI interpreting visual data
- speech recognitionMachines understanding human talk
- fraud detectionUncovering deceptive schemes
autonomous vehicles · chatbots · facial recognition · medical diagnosis · drug discovery · forecasting · content moderation · translation
Data used to teach models
- validationConfirming truth or accuracy
- cross-validationResampling method for model assessment
- overfittingExcess adaptation to samples
- underfittingModel too simple
accuracy · precision · recall · generalization · AUC · regularization · benchmark · ground truth
Machine learning computation library
- pandasAsian bamboo-eating bears
- GPUVisual processing hardware
- cloud computingInternet-delivered digital infrastructure
- data pipelineAutomated sequence processing information
Creating predictive input variables
- training dataInput for machine learning
- modelRepresentation or paradigm
- inferenceConclusion reached through reasoning
- predictionForecast of coming events
weights · parameters · embeddings · tokenization · corpus · dataset · annotation · epoch
neural networks
Layered computational learning systems
- deep learningLayered neural network methods
- reinforcement learningTraining through reward signals
- classificationSystematic grouping by traits
regression · clustering · decision trees · backpropagation · activation function · k-means
natural language processing
Computational analysis of speech
- NLPMachine language comprehension
- computer visionAI interpreting visual data
- speech recognitionMachines understanding human talk
fraud detection · autonomous vehicles · chatbots · facial recognition · medical diagnosis · drug discovery · forecasting · content moderation · translation · text generation
training set
Data used to teach models
- validationConfirming truth or accuracy
- cross-validationResampling method for model assessment
- overfittingExcess adaptation to samples
underfitting · accuracy · precision · recall · generalization · AUC · regularization · benchmark · ground truth · error rate
TensorFlow
Machine learning computation library
- pandasAsian bamboo-eating bears
- GPUVisual processing hardware
- cloud computingInternet-delivered digital infrastructure
feature engineering
Creating predictive input variables
- training dataInput for machine learning
- modelRepresentation or paradigm
- inferenceConclusion reached through reasoning
prediction · weights · parameters · embeddings · tokenization · corpus · dataset · annotation · epoch · convergence
Deriving specifics from generalities
a priori knowledge · logical inference · intuitive understanding · rationalism