구글MLB2024

뉴런은 입력을 받고 출력을 뱉는다.뉴런을 쌓으면 레이어가 된다.레이어가 모이면 뉴럴네트워크(NN, 인공신경망)으로 부른다.

Frobenius norm$||w^{l}||^2 = \\sum{i=1}^{n^{l}} \\sum{j=1}^{n^{l-1}} (w\_{i,j}^{l})^2$데이터셋 종류Train setDev set (Hold-out cross validation set)Test set결

ML strategyWhy Human-level Performance?Avoidable BiasUnderstanding Human-level PerformanceSurpassing Human-level PerformanceImproving your Model Perfo

CNNNeural Style Transfer

Recurrent Neural Networks (RNN) NLP & Word Embeddings Sequance to Sequence Architectures Speech Recognition Transfomers

An industrial accident analysis service based on precedents related to industrial accident cases.It aims to assist individuals who have been denied ap