Linganisha mbinu

	Uboreshaji wa Gradient Ulioimarishwa	LightGBM Iliyoimarishwa
Nyanja	Ujifunzaji wa Mashine	Ujifunzaji wa Mashine
Familia	Machine learning	Machine learning
Mwaka wa asili	2001 (gradient boosting); 2016 (explicit L1/L2 regularization in XGBoost)	2017
Mwanzilishi	Chen, T. & Guestrin, C. (building on Friedman, J. H.)	Ke, G. et al. (Microsoft Research)
Aina	Regularized ensemble (additive tree model)	Regularized gradient boosting ensemble
Chanzo asilia	Chen, T. & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794.	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.
Majina mbadala	penalized gradient boosting, shrinkage-regularized boosting, XGBoost-style regularization, L1/L2 gradient boosting	LightGBM with L1/L2 regularization, penalized LightGBM, LightGBM ridge/lasso, regularized LGBM
Zinazohusiana	6	5
Muhtasari	Regularized gradient boosting extends the classic additive tree ensemble (Friedman 2001) by embedding L1 and L2 penalty terms directly into the training objective, along with a complexity penalty on tree size. Popularized by XGBoost (Chen & Guestrin 2016), this framework reduces overfitting and improves generalization compared to unpenalized boosting, while retaining the method's characteristic accuracy on tabular data.	Regularized LightGBM applies L1 (lasso) and L2 (ridge) penalty terms to the leaf weight objective of LightGBM — Microsoft's highly efficient gradient boosting framework — to control model complexity, reduce overfitting, and improve generalization on tabular classification and regression tasks with high-dimensional or noisy feature sets.