önszupervizált

Önszupervizált is a term of Hungarian origin that translates to "self-supervised" in English. It refers to a type of machine learning where the model learns from unlabeled data by creating its own supervisory signals. Instead of relying on human-provided labels, the algorithm is tasked with solving a "pretext" task. For example, a model might be given an image with a patch missing and its task is to predict the missing patch. By successfully completing these pretext tasks, the model learns useful representations of the data, such as understanding visual features or grammatical structures.

The core idea behind önszupervizált learning is that the inherent structure of the data itself can provide sufficient supervision. This allows for the training of powerful models on vast amounts of readily available unlabeled data, which is far more abundant than labeled data. Once trained on these pretext tasks, the model's learned representations can then be fine-tuned for downstream tasks, such as image classification or natural language understanding, with a much smaller amount of labeled data. This approach has significantly advanced the field of artificial intelligence, particularly in areas like computer vision and natural language processing, by reducing the dependency on costly and time-consuming manual data annotation.