Please credit the link and author of this article when reprinting, thank you for your cooperation!

Micro Signal: Machine Learning Bee

This is the final lesson in the Cornerstones of Machine Learning. The main focus is on three very important and extremely easy to overlook principles in machine learning.

Three principles of machine learning

Occam’s razor

entia non sunt multiplicanda praeter necessitatem

(entities must not be multipliedbeyond necessity)—William of Occam (1287-1347)

Corresponding to machine learning, theSimple model first (linear first)

Sampling Bias

If the data is sampled in a biased way, learning will produce a similarity biased outcome

Sampling bias, especially at validation, should be avoided.

Data Snooping

If a data set has affected any step in the learning process, its ability to assess the outcome has been com- promised.

It is especially important to avoid the human eye prying into the data.

careful balance between data-driven modeling (snooping) and validation (no-snooping)

Next

summary

This post introduces three important principles in machine learning.

This concludes the Machine Learning Cornerstones class and will be followed by a hands-on machine learning system, so stay tuned!

Finally, thank you from the bottom of my heart, Mr. Lin, for watching this course and being enlightened.

Reference.

1. Screenshot of this article from Mr. Lin Tianxuan's class

Thank you for your interest in Machine Learning Bee, and I look forward to your comments and suggestions.

Recommended>>
1、Investment Company Successfully Completes 2018 Network and Information Security Inspections
2、Section 7 of the vue video tutorial series How to bind events to a tag on a template
3、Top Gun the future of blockchain building
4、Top 10 cases of intelligent applications in steel mills
5、Why internet startup opportunities will become less available in 2018

已发送

发送中