Generalisation refers to the ability of an artificial intelligence or machine learning model to apply its learning to new situations or data that are not in its original training set. In other words, generalisation implies that a model is able to learn general patterns and characteristics of a dataset and apply that knowledge to new data.
Generalisation is a fundamental aspect of machine learning, since the goal of training a model is to enable it to make accurate predictions on data it has never seen before. If a model is only able to make accurate predictions on the data used to train it, it is said to have overfitted or memorised the training set.
The generalisability of a model can be improved by various techniques, such as regularisation, cross-validation, feature selection and collecting more training data. In general, the larger and more diverse the training data set, the better the generalisability of the model.
Fernando Pavón, CEO of Gamco and expert in Artificial Intelligence applied to business explains to us in the AceleraPYMES cycle how small companies can [...]
Read More »The semantic web or "internet of knowledge" is an extension of the current web. Unlike the latter, the semantic web is based on proportional [...]
Read More »Collecting debts, nowadays, is becoming an arduous task for many companies or freelancers. More and more banks, debt collection [...]
Read More »Once the basic concepts for building a commercial software with artificial intelligence are clear, where it is defined to whom to dedicate effort and [...]
Read More »
