deep learning :the effect of pooling
鏈接：https://www.zhihu.com/question/36686900/answer/130890492 來源：知乎 主要是兩個作用： 1. invariance(不變性)，這種不變性包括translation(平移)，rotation(旋轉)，scale(尺度) 2. 保留主要的特徵同時減少參數(降維，效果類似PCA)和計算量，防止過擬合，提高模型泛化能力 (1) translat
相關文章
- 1. Review of Semantic Segmentation with Deep Learning
- 2. Deep Learning State of the Art (2020) | MIT Deep Learning Series記錄
- 3. Deep learning：二十三(Convolution和Pooling練習)
- 4. [paper]The Limitations of Deep Learning in Adversarial Settings(JSMA）
- 5. Exploring the teaching of deep learning in neural networks
- 6. The Limitations of Deep Learning in Adversarial Settings
- 7. Neuroevolution: A different kind of deep learning
- 8. Inspirational Applications of Deep Learning
- 9. Evaluation of Deep Learning Toolkits
- 10. Securing the Deep Learning Stack
- 更多相關文章...
- • XSLT
- • XSLT
- • Java Agent入門實戰（一）-Instrumentation介紹與使用
- • Java Agent入門實戰（三）-JVM Attach原理與使用
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. Review of Semantic Segmentation with Deep Learning
- 2. Deep Learning State of the Art (2020) | MIT Deep Learning Series記錄
- 3. Deep learning：二十三(Convolution和Pooling練習)
- 4. [paper]The Limitations of Deep Learning in Adversarial Settings(JSMA）
- 5. Exploring the teaching of deep learning in neural networks
- 6. The Limitations of Deep Learning in Adversarial Settings
- 7. Neuroevolution: A different kind of deep learning
- 8. Inspirational Applications of Deep Learning
- 9. Evaluation of Deep Learning Toolkits
- 10. Securing the Deep Learning Stack