Deep Learning

Deep learning is a sub-area of machine learning based on artificial neural networks and allows computers to learn from data on their own.

The aim is to reproduce the functioning of the human brain by using several layers of artificial neural networks.

Deep learning consists of specially designed models with multiple layers of neurons. At the beginning are the input neurons, which receive the input data. At the end are the output neurons, which provide classifications or predictions, for example. The results are usually output as probabilities.
Between these are several hidden layers that process the data step by step. In this process, features are automatically recognized and represented in increasingly abstract terms. The more layers a model has, the more complex relationships it can capture.

Large amounts of data and complex training methods can be used to develop models with high predictive accuracy. These are e.g. used in image or speech recognition.

Examples

• App automatically detects faces on photos
• Voice assistant understands and processes spoken commands
• Translation tool translates texts into different languages
• Streaming platform recommends content based on user behavior