知识卡片
对抗生成同时训练判别
生成对抗网络让生成器产生样本,让判别器区分生成与真实样本,再通过双方目标更新参数。它能建模数据分布,并不保证生成样本真实或对推荐有用。具体推荐方案必须说明生成对象、训练信号与验收指标。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- · 查看原书依据
The generating model is trained to create new examples, and the discriminating model classifies examples as either real or counterfeit. GAN generates data from scratch, primarily images and music. The idea here is to create new examples from existing distribution. The training strategy used defines a game between two competing networks. A minimax function is formulated based on data and model distribution such that the
参考来源
- 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第03章《深度推荐与评估指标》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications/_epub-src/OEBPS/Text/9789815136746-C3.html)
- 结论依据:摘录中关于对抗生成同时训练判别的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:The generating model is trained to create new examples, and the discriminating model classifies examples as either real or counterfeit. GAN generates data from scratch, primarily images and music. The idea here is to create new examples from existing distribution. The training strategy used defines a game between two competing networks. A minimax function is formulated based on data and model distribution such that the