Idea Transcript
What is feature hashing (hashing-trick)? We have 83 open jobs ª
Imagine yourself at Comcast
Learn more
I know feature hashing (hashing-trick) is used to reduce the dimensionality and handle sparsity of bit vectors but I don't understand how it really works. Can anyone explain this to me.Is there any python library available to do feature hashing? Thank you. python hash vector machine-learning edited Dec 30 '11 at 23:18
maxy 2,391
15
18
asked Dec 29 '11 at 20:29
Maggie 1,886
5
27
48
Are you looking for something like this? shogun-toolbox.org – S.Lott Dec 29 '11 at 20:32
3 Answers
On Pandas, you could use something like this: import pandas as pd import numpy as np the quick brown fox" transform to: h(the) mod 5 = 0 h(quick) mod 5 = 1 h(brown) mod 5 = 1 h(fox) mod 5 = 3
Use index rather then text value, saves space. To summarize some of the applications: dimensionality reduction for high dimension feature vector text in email classification task, collaborate filtering on spam sparsification bag-of-words on the fly cross-product features multi-task learning Reference: Origin paper: 1. Feature Hashing for Large Scale Multitask Learning 2. Shi, Q., Petterson, J., Dror, G., Langford, J., Smola, A., Strehl, A., & Vishwanathan, V. (2009). Hash kernels What is the hashing trick Quora Gionis, A., Indyk, P., & Motwani, R. (1999). Similarity search in high dimensions via hashing Implementation: Langford, J., Li, L., & Strehl, A. (2007). Vow- pal wabbit online learning project (Technical Report). http://hunch.net/?p=309. edited Feb 24 '17 at 5:36
answered Nov 7 '15 at 10:08
CodeFarmer 1,428
14
22
Can you comment on the impact of feature hashing on the learned model? since there will be hash collisions. Yes I know they're improbable and minimal etc, but collisions will occur; what is the impact of these collisions on the learned model? any pointers to research that looks into this question is appreciated. One thing is clear, the learned model from hashed features is NOT guaranteed to be the same Model you get from the original un-hashed features. How do they differ and to what degree? – Kai May 21 '16 at 17:03
@Kai I'v added original paper on this topic. Error boundary was analysed so does empirical results. Please take a look. – CodeFarmer Feb 24 '17 at 5:40
much appreciated, thank you. – Kai Feb 24 '17 at 18:08
Join Stack Overflow to learn, share knowledge, and build your career. Email Sign Up OR SIGN IN WITH
Google Facebook
Learn more