yzhao062/pyod

AutoEncoder does not generate a symmetric model

Open

#138 创建于 2019年11月3日

在 GitHub 查看
 (3 评论) (0 反应) (0 负责人)Python (7,762 star) (1,308 fork)batch import
bughelp wanted

描述

Hi,

I plotted the model when printing the model summary: https://github.com/yzhao062/pyod/blob/da68f50ee79bd74777d2d406cac6dcfad412e599/pyod/models/auto_encoder.py#L187

And it seems that you added an extra layer at the beginning with the same dimension as the input. I don't know why this is required because it would generate an asymmetric model. Here is the model I printed out:

model

The final dimension would be like 645-645-100-100-500-9-500-100-100-645.

I think the problem might be caused by this line: https://github.com/yzhao062/pyod/blob/da68f50ee79bd74777d2d406cac6dcfad412e599/pyod/models/auto_encoder.py#L224

The performance is similar after I commented out the above line.

Please correct me if I miss something! Thanks!

贡献者指南

技术栈
pythontensorflow
领域
machine learning
议题类型
bug
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
2
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
under 1 hour
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
clear
前置要求
Basic knowledge of autoencodersPython programming
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
80
研究方向
Investigate the autoencoder model construction in pyod/models/auto encoder.py, specifically around lines 187 and 224. The issue suggests that an extra layer with the same dimension as input causes asymmetry. The fix may be to remove or adjust that layer. Review the model summary and confirm the symmetry after modification.