목록Total (86)
Head vs breakz
[arxiv] - Machine Learning in Python: Main developments and technology trends in data science, machine learning, and artificial intelligence
https://arxiv.org/abs/2002.04803?context=cs Machine Learning in Python: Main developments and technology trends in data science, machine learning, and artificial intelligen Smarter applications are making better use of the insights gleaned from data, having an impact on every industry and research discipline. At the core of this revolution lies the tools and the methods that are driving it, from..
[arxiv] - On Feature Normalization and Data Augmentation
https://arxiv.org/abs/2002.11102v1 On Feature Normalization and Data Augmentation Modern neural network training relies heavily on data augmentation for improved generalization. After the initial success of label-preserving augmentations, there has been a recent surge of interest in label-perturbing approaches, which combine features an arxiv.org 현대 신경망 훈련은 일반화를 위해 데이터 확대에 더욱 의존적이다. label-preser..
[arxiv] - A Primer in BERTology: What we know about how BERT works
https://arxiv.org/abs/2002.12327v1 A Primer in BERTology: What we know about how BERT works Transformer-based models are now widely used in NLP, but we still do not understand a lot about their inner workings. This paper describes what is known to date about the famous BERT model (Devlin et al. 2019), synthesizing over 40 analysis studies. We als arxiv.org Transformer-based models 모델은 NLP에서 많이 사..
[인공지능]- One shot learning with Siamese Networks using keras - Harshall lamba - 3
One shot learning with Siamese Networks using keras - Harshall lamba https://towardsdatascience.com/one-shot-learning-with-siamese-networks-using-keras-17f34e75bb3d 를 공부하기 위해 번역과 정리 하는 글입니다. Base Line 1 — Nearest Neighbor Model 간단한 baseline 모델을 만들고 그 결과를 만들려고하는 복잡한 모델과 비교하는 것이 좋다. 첫번째 기본 모델은 Nearest Neighbor model이다. 벡터 X와 다른 벡터들 간의 L2 distance를 비교하여, 거리가 가장 작은 벡터를 확인한다. 거리가 가까울수록 유사성이 높기 떄문에, 그..