Numerical Training Dataset
A Numerical Training Dataset is a Training Dataset which both, independent variable and target ouput have numerical values.
- Example(s)
- Counter-Example(s):
- See: Training Database, Training Record, Cross-Validation Task, Classifier, Overfitting, Dataset, Supervised Machine Learning System, Machine Learning.
References
2018
- (Wikipedia, 2018) ⇒ https://en.wikipedia.org/wiki/Training,_test,_and_validation_sets#Training_dataset Retrieved:2018-4-1.
- A training dataset is a dataset of examples used for learning, that is to fit the parameters (e.g., weights) of, for example, a classifier.[1] [2]
Most approaches that search through training data for empirical relationships tend to overfit the data, meaning that they can identify apparent relationships in the training data that do not hold in general.
- A training dataset is a dataset of examples used for learning, that is to fit the parameters (e.g., weights) of, for example, a classifier.[1] [2]
- ↑ Ripley, B.D. (1996) Pattern Recognition and Neural Networks, Cambridge: Cambridge University Press, p. 354
- ↑ "Subject: What are the population, sample, training set, design set, validation set, and test set?", Neural Network FAQ, part 1 of 7: Introduction (txt), comp.ai.neural-nets, Sarle, W.S., ed. (1997, last modified 2002-05-17)
2017
- (Sammut & Webb, 2017) ⇒ (2017) "Training Set". In: Sammut, C., Webb, G.I. (eds) "Encyclopedia of Machine Learning and Data Mining". Springer, Boston, MA
- QUOTE: A training set is a data set containing data that are used for learning by a learning system. A training set may be divided further into a growing set and a pruning set.