閱讀人工智能論文的十條建議

（給機(jī)器學(xué)習(xí)算法與Python實(shí)戰(zhàn)加星標(biāo)，提升AI技能）

1.GloVe (2014)

• 論文鏈接：https://www.aclweb.org/anthology/D14-1162.pdf

• Transformers：http://papers.nips.cc/paper/7181-attention-is-all-you-need

• Word2Vec:https://arxiv.org/abs/1301.3781
• BERT:https://arxiv.org/abs/1810.04805

2.AdaBoost (1997)

• 論文鏈接：https://www.sciencedirect.com/science/article/pii/S002200009791504X

• 隨機(jī)森林分類器：https://en.wikipedia.org/wiki/Random_forest
• 梯度提升技術(shù)：https://en.wikipedia.org/wiki/Gradient_boosting
• XGBoost軟件包：https://github.com/dmlc/xgboost
• LightGBM：https://github.com/microsoft/LightGBM

3.Capsule Networks (2017)

• 論文鏈接：https://arxiv.org/abs/1710.09829

4.Relational Inductive Biases (2018)

• 論文鏈接：https://arxiv.org/pdf/1806.01261.pdf

• 2020年圖機(jī)學(xué)習(xí)的主要趨勢：https://towardsdatascience.com/top-trends-of-graph-machine-learning-in-2020-1194175351a3

5.Training Batch Norm and Only BatchNorm (2020)

• 論文鏈接：https://arxiv.org/abs/2003.00152

• 彩票假說：https://arxiv.org/abs/1803.03635

6.Spectral Norm (2018)

• 論文鏈接：https://arxiv.org/abs/1802.05957

• Wasserstein損失：https://arxiv.org/abs/1701.07875

• dropout:https://en.wikipedia.org/wiki/Dropout_(neural_networks)

7.Perceptual Losses (2016)

• https://medium.com/ml-cheat-sheet/winning-at-loss-functions-2-important-loss-functions-in-computer-vision-b2b9d293e15a

8.Nadam (2016)

• Radam：https://arxiv.org/abs/1908.03265v1
• Lookahead：https://arxiv.org/abs/1907.08610
• Ranger：https://github.com/lessw2020/Ranger-Deep-Learning-Optimizer

9.The Double Descent Hypothesis (2019)

• 圖像分類的“技巧包”：https://arxiv.org/abs/1812.01187

10.On The Measure of Intelligence (2019)

• https://arxiv.org/abs/1911.01547

• https://youtu.be/UX8OubxsY8w