co-trained

English

Etymology

From co- +‎ trained.

Pronunciation

  • IPA(key): /ˌkoʊˈtɹeɪnd/

Adjective

co-trained (not comparable)

  1. (artificial intelligence, machine learning) Trained jointly or simultaneously alongside another model, classifier, or network component, typically processing distinct feature sets, data views, or modalities.
    • 2016, S. Bianco et al., CURL: Image Classification using co-training..., ScienceDirect:
      ...co-trained classifiers (C). The difference between the two image representations is that one is built on the combination of all the image features...
    • 2021, X. Huang et al., "Deep Domain Adaptation based Cloud Type Detection...", NASA Technical Reports Server (NTRS):
      The l2 loss is co-trained with the correlation alignment and classifier losses in an end to end fashion...
    • 2026 August 6, Google DeepMind, WeatherNext: AI model achieves breakthrough in forecasting cyclones:
      The model was co-trained on two distinct data modalities: global weather dynamics and expert-curated historical cyclone observations.
  2. (artificial intelligence, machine learning, by extension) Mutually optimized or enhanced through a collaborative machine learning architecture, such that individual components iteratively swap labeled data, align feature distributions, or act as a corrective ensemble to boost overall predictive accuracy.
    • 2019, A. F. da Costa et al., "Boosting collaborative filtering with an ensemble...", ScienceDirect:
      In order to combine and improve the individual models generated by CoRec as they are simultaneously co-trained, ECoRec incorporates an in-built ensemble...
    • 2022, H. Lang et al., “Co-training Improves Prompt-based Learning for Large Language Models”, in Proceedings of Machine Learning Research:
      For CB, the co-trained label model outperforms GPT-3 32-shot despite only using 4 labeled examples.
    • 2025 July 16, NVIDIA Research, R²D²: Training Generalist Robots with NVIDIA Research:
      Even with 400 real demonstrations, the co-trained policy consistently outperforms the real-only policy by an average of 38%...
  • co-train
  • co-training