Recently, many friends have asked me how to get started with machine learning and big data. In order to answer everyone's doubts, I wrote this article for this purpose, hoping to help everyone. There is a lot of learning material at the end of the article.
Getting Started with Machine Learning
Machine learning is a multi-disciplinary science, and the requirements for beginners are relatively high. The basic requirements are as follows.
(1) Mathematics: Advanced Mathematics + Linear Algebra + Probability Theory. These courses can be learned first, and then gradually deepened when learning machine learning encounters specific knowledge.
(2) Programming: Machine learning has certain requirements for programming, and the programming language is preferably Python, because there are many materials in this area.
After you have the above foundation, you can watch some basic video courses, such as the machine learning course taught by Stanford University Andrew ng in the NetEase Open Class , which is recognized as the best basic video. There is also a video course of Professor Li Hongyi of National Taiwan University on Bilibili. The editor thinks that this course is still very interesting. Professor Li Hongyi's lecture is very interesting, and the course focuses on practice.
After you have an overall understanding of machine learning, you can read Li Hang's "Statistical Learning Methods" in order to improve the theory, and you can read "Machine Learning in Practice" in order to improve the practice. The PDF of these materials has a download link at the end of this article.
Finally, you can do some practical projects, such as participating in the Kaggle competition, or you can read "Recommendation System Practice". Continuing to study requires more practice and more reading of papers.
Getting Started with Big Data
The editor thinks that the threshold for learning big data is much lower than machine learning. As long as you have a foundation in JAVA language, you can learn big data. But if you want to have some attainments in big data, I think it is also very difficult.
The current mainstream big data processing frameworks are Hadoop and Spark. Hadoop is mainly used in batch computing, and Spark is the integration of multiple computing modes. I think Spark should be the mainstream in the future, but Hadoop is currently the mainstream.
To learn Hadoop, you can watch the public course video of Tanzhou College. This article has a video link to Baidu Cloud. The PDF materials include "In-depth Understanding of Big Data Big Data Processing and Programming Practice", "Hadoop Technology Insider" and so on.
Learning big data must have project experience, because only in the project can a deep understanding of big data. For in-depth study, you can look at the source code of Hadoop and Spark, and see more papers.
learning materials
Machine Learning PDFs
百度云链接：https://pan.baidu.com/s/1sTnXrKQAZR3VoaqFrAO9gA
密码：pdtz
大数据视频资料
百度云链接：https://pan.baidu.com/s/1EOZsWW1Pp1ospEPVRiF0GQ
密码：d25d
大数据PDF资料
百度云链接：https://pan.baidu.com/s/161Ckvx30vBcGEczLPGEFiA
密码：jjt7
获取更多干货请关注微信公众号：追梦程序员。
