training set
Model's teaching data
A training set is the collection of data used to teach a machine learning model.
The algorithm learns patterns from this curated dataset, adjusting its parameters to make accurate predictions.
Training sets are distinguished from test sets, which evaluate how well the model performs on data it hasn't seen: We split the data 80/20 between training set and test set.
- datasetOrganized collection of information
- validationConfirming truth or accuracy
- labelsIdentifying markers attached to items
- modelRepresentation or paradigm
- featuresDistinctive qualities or aspects
feature engineering · samples · overfitting · neural network · algorithm · machine learning · examples · supervised · annotations · corpus · batch · epoch · learning · classifier · data points · ground truth · split · preprocessing