Not much to say, I believe that you who read this article must be interested in data analysis, data mining, or want to engage in and aspects. This article will no longer repeat the importance of python to data analysis, the origin of data analysis and so on.
Here, I will go straight to the point, and I will tell you how to learn from my experience of learning data mining for 3 years, so that everyone can avoid detours. Pure personal opinion, if there is something wrong, please leave a message for advice.
Not much to say, just put the picture directly.
One learn the tools
python language
It is recommended to read Liao Xuefeng's python3 tutorial .
Data analysis python basics
Such as list, tuple, dic, set, etc. I will write about it in my future blog.
Two get data
python crawler
Recommend a book: "Web Scraping with Python" written by Ryan Mitchell, very good. After reading this book + combating reptiles a few times, you will be proficient. The crawler blog I will write in the future will also be written with reading notes when reading this book.
I recommend Cui Qingcai's blog in actual combat , and you can also read the crawler combat blog I wrote later. I studied according to his big framework.
Three data storage and reading
IO, EXCEL, CSV, JSON, SQL database of data. HDF5, etc.
四 NUMPY PANDAS SCIPY MATPLOTLIB
Numpy array
Pandas data analysis
Scipy matrix
Matplotlib data visualization
This part recommends reading "Python Data Analysis" (Python Data Analysis) written by Lvan Idris.
Five data preprocessing
Preprocessing of collected or ready-made data, data cleaning (recommended to see "Clean Data = Clean Data: Introduction and Practice of Data Cleaning" by Megan Squire), data integration, data transformation, data reduction, etc. If you are interested in this part, you can read a book "Python Data Analysis and Mining Practice". Just look at the framework of this book. It's actually not well written. Wasted my time.
Six Modeling Machine Learning
Learn various machine learning, data analysis algorithms.
The algorithm principle is recommended to see "Top Ten Data Mining Algorithms" Author: Xindong Wu, Vipin Kumar
Python implementation is recommended to see "Python Data Analysis and Mining Practice" + "Python Data Analysis" + "Machine Learning Practice"
Seven Neural Networks
Further down is deep learning, neural networks.
Message:
Learning each one takes a lot of time, energy and effort. I hope you are really interested in data mining and data analysis.