yzhao062/pyod

LOCI too much memory and run time

Open

#106 创建于 2019年6月7日

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

描述

I don't know why this is happening but the training time for LOCI implementation (fit) take too much time and memory for even a10,000X30 data, please try to run it for credit card fraud detection kaggle dataset

贡献者指南

技术栈
python
领域
performancemachine learning
议题类型
performance
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。
4
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。
1-2 days
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。
needs investigation
前置要求
Basic understanding of LOCI algorithmFamiliarity with Python profiling
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。
25
研究方向
Investigate the LOCI implementation in pyod (likely in pyod/models/loci.py). Profile memory and time usage using Python profiling tools. Check if the algorithm can be vectorized or optimized. Review comments in the issue for maintainer suggestions (there are 3 comments and the issue is assigned).