Stochastic Gradient Descent (SGD)

Concept and definition

Stochastic Gradient Descent (SGD)

What is Stochastic Gradient Descent (SGD)?

Stochastic Gradient Descent (SGD) is an optimisation algorithm used in training machine learning models.

The main idea behind stochastic gradient descent is to minimise a cost function by iteratively adjusting model parameters based on feedback from the training data set. Instead of updating the model parameters exhaustively with all the training data at each iteration (as in regular gradient descent), stochastic gradient descent uses small random samples of the training data (known as "minibatches" or "batches") at each iteration to update the model parameters.

Stochastic gradient descent is especially useful when the training data set is very large, as it allows the model to train more efficiently by processing small samples of the data at a time. In addition, stochastic gradient descent can help prevent the model from becoming trapped in local optima and converge more quickly to a global optimum.

Stochastic gradient descent is a widely used algorithm in the optimisation of deep learning models, and is essential for training large neural networks that require large training data sets.

« Back to glossary

Do you want to get in touch?

CDRs contain data that a telecommunications company collects about phone calls, such as time and length of call. This data can be used in analytical applications.
Fill the form
Share:
Big Data applied to business

The Big Data market is booming. Although the need to transform data into information for decision making is not new, the need to [...]

Read More »
How to detect delinquent customers and avoid defaults? 10 signs of delinquency

As a consequence of this pandemic and economic situation in which we have found ourselves for the last two years, with the intention of better protecting the [...]

Read More »
Artificial intelligence applied to the semantic web 

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 »
AI against non-payments and defaulters

An article published in April 2021 by Óscar Jiménez El Confidencial, was titled "34,000 M prize for banks for applying well i [...]

Read More »
See more entries
© Gamco 2021, All Rights Reserved - Legal notice - Privacy - Cookies