Machine learning · BEGINNER LESSON
Training set
A training set is the part of a data set used to train a machine learning (ML) model. In supervised learning, it contains examples with input features and known target values or labels that the model uses to learn patterns.
What it is
A training set is the subset of a data set that is used for training a machine learning (ML) model. In supervised learning, the training set contains examples made of input features and known target values or labels. Features are measured properties of each sample, and the label is the outcome the model should learn to predict. The goal is for the model to find a mapping from inputs to outputs by adjusting its internal parameters during training. [0,1,2]
[1] pmc.ncbi.nlm.nih.gov · articles · PMC11980903Machine learning for microbiologists - PMC [2] 6.1 What Is Machine Learning? - Principles of Data Science [3] Trade-off between training and testing ratio in machine ... - PMC
How it is used
During training, the algorithm looks for patterns and relationships in the training set and updates the model step by step. After training, the model is checked on separate data, often called a test set or validation set, to see whether it can predict new cases well. Keeping the training set separate from evaluation data helps show whether the model learned useful structure instead of just memorizing the training examples. [0,1,2]
[1] pmc.ncbi.nlm.nih.gov · articles · PMC11980903Machine learning for microbiologists - PMC [2] 6.1 What Is Machine Learning? - Principles of Data Science [3] Trade-off between training and testing ratio in machine ... - PMC
Limits and common problems
A training set must be large and good enough for the task. If the data are poor quality, the resulting predictions may be unreliable; this is the idea behind garbage in, garbage out. If the training set is too small or the model is too rigid, the model may underfit and miss important patterns. If it is too tightly fitted to the training set, it may overfit and perform worse on new data. Sources also note that performance on cross-validation or one data set may not carry over to unseen data from another data set. [0,1,2]
[1] pmc.ncbi.nlm.nih.gov · articles · PMC11980903Machine learning for microbiologists - PMC [2] 6.1 What Is Machine Learning? - Principles of Data Science [3] Trade-off between training and testing ratio in machine ... - PMC
SEE THE IDEA
How a training set fits into machine learning
Examples with features and known labels used to teach the model
The algorithm adjusts parameters to learn patterns
Separate data used to check performance
Unseen cases the trained model will predict
The training set teaches the model first; separate evaluation data checks whether the learned pattern works beyond that set.
A training set is for learning, not final evaluation.
TRY IT YOURSELF
Test your understanding
Choose an answer, then check the explanation.
PUT IT TOGETHER
Match each concept to its meaning
Drag a label onto its description, or select a label and then select a description. This also works with a keyboard.
4 pairs to match
Sources and further reading
This general lesson was generated from public source material. It explains a concept; it does not interpret a private paper or provide medical advice.