পদ্ধতির তুলনা করুন

	নলেজ ডিস্টিলেশন (Knowledge Distillation)	নিউরাল আর্কিটেকচার সার্চ	সাপোর্ট ভেক্টর মেশিন (শ্রেণীকরণ)
ক্ষেত্র	গভীর শিখন	গভীর শিখন	যন্ত্র শিখন
পরিবার	Machine learning	Machine learning	Machine learning
উদ্ভবের বছর	2015	2017	1995
প্রবর্তক	Hinton, G., Vinyals, O. & Dean, J.	Zoph, B. & Le, Q.V.	Cortes, C. & Vapnik, V.
ধরন	Neural network compression (teacher–student)	Automated architecture optimization (deep learning)	Maximum-margin classifier (kernel method)
মৌলিক উৎস	Hinton, G., Vinyals, O. & Dean, J. (2015). Distilling the Knowledge in a Neural Network. NeurIPS Deep Learning Workshop.	Zoph, B. & Le, Q.V. (2017). Neural Architecture Search with Reinforcement Learning. ICLR.	Cortes, C. & Vapnik, V. (1995). Support-Vector Networks. Machine Learning, 20, 273–297.
অপর নাম	Bilgi Damıtma (Knowledge Distillation), bilgi damıtma, teacher-student distillation, model distillation	Nöral Mimari Arama (NAS), NAS, automated architecture design, differentiable architecture search	Destek Vektör Makinesi (SVM — Sınıflandırma), support-vector network, SVM classifier, maximum-margin classifier
সম্পর্কিত	5	5	5
সারসংক্ষেপ	Knowledge Distillation is a model-compression technique, introduced by Geoffrey Hinton and colleagues in 2015, that trains a small student model using the soft-label outputs of a large teacher model. Distilled models such as DistilBERT and TinyBERT reach roughly 97% of the larger model's performance while running far faster.	Neural Architecture Search (NAS), introduced by Zoph and Le in 2017, automatically optimizes architectural decisions such as a network's depth, width, and connection structure instead of hand-designing them. Leading methods in the field include DARTS, ENAS, and Once-for-All.	The Support Vector Machine, introduced by Corinna Cortes and Vladimir Vapnik in 1995, is a classifier that finds the optimal separating hyperplane between classes in a high-dimensional space. It chooses the boundary that leaves the widest possible margin to the nearest training points, which makes its decisions robust on new data.