Porovnat metody

	Self-supervised LightGBM	Semi-supervised LightGBM
Obor	Strojové učení	Strojové učení
Rodina	Machine learning	Machine learning
Rok vzniku	2017–2020	2017–2019
Tvůrce	Ke, G. et al. (LightGBM); self-supervised paradigm adapted from broader SSL literature	Ke, G. et al. (LightGBM); semi-supervised extension via community practice and research
Typ	Hybrid (self-supervised pretraining + gradient boosting)	Semi-supervised gradient boosting ensemble
Původní zdroj	Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: A Highly Efficient Gradient Boosting Decision Tree. Advances in Neural Information Processing Systems, 30.	Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: A highly efficient gradient boosting decision tree. Advances in Neural Information Processing Systems, 30, 3146–3154.
Další názvy	SSL-LightGBM, self-supervised gradient boosting, pretraining LightGBM, pseudo-label LightGBM	SSL-LightGBM, pseudo-label LightGBM, self-training LightGBM, semi-supervised GBDT
Příbuzné	6	4
Shrnutí	Self-supervised LightGBM combines the self-supervised learning paradigm with the LightGBM gradient boosting framework to exploit large volumes of unlabeled tabular data. A self-supervised pretext task — such as masked feature prediction or contrastive corruption — generates rich feature representations or pseudo-labels that are then used to train or fine-tune a LightGBM model, substantially improving performance in label-scarce regimes.	Semi-supervised LightGBM combines LightGBM's highly efficient gradient boosting framework with semi-supervised strategies — most commonly pseudo-labeling or self-training — to exploit large pools of unlabeled data alongside a smaller labeled set, improving predictive performance when obtaining labels is costly or time-consuming.