Katastrophales Vergessen is a phenomenon observed in künstliche Intelligenz, particularly in maschinellem Lernen models, where a model forgets previously learned information upon learning neue Daten or tasks. This issue is especially prominent in neural networks that are trained sequentially on different tasks, leading to a decline in performance on earlier tasks.
Der Begriff wurde erstmals im Zusammenhang mit psychologischer Lerntheorie beschrieben, but it has significant implications in AI and machine learning. When a model is trained on one task and then retrained on a different task, it often fails to retain the knowledge acquired from the first task, resulting in poor performance on it. This is problematic for applications where continuous learning and adaptation are required.
Catastrophic forgetting is commonly attributed to the way neural networks update their weights during training. When new data is introduced, the adjustments made to the network’s parameters can interfere with the previously established representations, causing them to degrade or be overwritten. As a result, the model exhibits a high degree of sensitivity to the order in which tasks are presented.
Um katastrophales Vergessen zu mildern, haben Forscher verschiedene Techniken erforscht, darunter:
- Regularisierung Methoden: These help maintain the weights associated with previous tasks while allowing new learning.
- Speicher-ergänzte Netzwerke: These incorporate externe Speicher Strukturen zur Speicherung von Informationen aus vorherigen Aufgaben.
- Progressive neuronale Netzwerke: These architectures allocate separate resources for each task, preserving knowledge from earlier tasks.
Das Verständnis und die Bewältigung des katastrophalen Vergessens sind entscheidend für die Entwicklung KI-Systemen that can learn continuously and adapt to new information without losing valuable knowledge.