Went to the weekend, it's time with a new wave! People often ask them, how I switch to the data analysis of the industry, but also how python into the pit from zero programming, the face of so many on the market "xx days of entry" books, how to choose? Today the small text will stroke a stroke of my school (ru) Xi (keng) path and method.
A book holds a house of gold, a beautiful woman in the book, that is true, but only if you have to find the book house of gold, plus independent thinking, and strive to practice. After learning methods summarized a number of big brother, little text to find the most efficient path to the entry that is selected direction, look at a book, more than a few yards of code! The important thing still needs to be said three times, and choose the right direction, look at a book, more than a few yards of code! Choose the right direction, look at a book, more than a few yards of code! Choose the right direction, look at a book, more than a few yards of code! And my direction data analysis, data analysis, so this article will enlighten - Basic - Advanced route described.
-----------------------
Into the pit first book, data analysis enlightenment book --- "layman Data Analysis"
"Data analysis in layman's language" is one of a series in layman's language, vivid language, describes the scene from various application methods and data analysis, is a fascinating analysis of the data enlightenment books. What impressed me the most profound way to Bayesian statistics and subjective probability these chapters, especially subjective probability, let the experts to an event score to predict the probability of an event, then according to the latest news, by Bayes' rule to adjust the probability value, re-forecasting.
-----------------------
With interest began after laying the foundation, because the family felt pretty good read layman's language, so the "easy to understand statistics", "in layman's language mysql", "in layman's language python" read over and over, for statistics, mysql, python have a a certain understanding.
Statistics addition to the "easy to understand statistics", as well as some of the more interesting books to help easy ride statistics, and that is to learn comic series "comic Statistics", "return to the comics statistical analysis", "comic Statistics Studies of factor analysis " , " comic linear Algebra "," comic calculus "and" comic differential equations " . These books are relatively easy to understand, so not much time spent.
Mysql main saw "layman mysql" and "mysql must know will be" basically applied aspects of the operation to cross the border.
About python, then classic "use python for data analysis" is recognized as a must-see teaching materials, I also saw, but I personally feel that is not the most important, the most important is my "stupid methodology python3" , this how books define it? Is that we can not know what it is talking about, but have to go along with its code books, a few yards from start to finish, even if the basic grammar basic python cross the border, is so magical, of course, still have to take the attitude of learning, rather than no brain code code.
Then "Contrast excel, easy to learn python Data Analysis" , known as the small yellow book, recommended small yellow book because it starts with the familiar excel, compare excel common operations to learn python can take you quickly master python data analysis for common operations, is a very easy to use and easy to read simple books, followed by the code several times, the basic pandas, numpy, matplotlib usage are well aware of.
Next is the "use python for data analysis" , compared to the small yellow book, this classic more in-depth detailed description of how to use the common package python, still very readable.
Finally, and most important point is to see official documents! The important thing to say n many times, see the official document! See the official documents! See the official documents! ... Official documents! ... party documents! ... Document! …files!
Whether a novice or an expert, you often need to look at official documents, after all, there are many ways a package, there are many parameters of each method, you want to remember eleven unlikely, but still commonly used to remember!
-----------------------
After lay the foundation, already qualified for simple data analysis, the next step is to expand our data analysis thinking, I recommend "utilization data - driven business analysis of real data" , which is a departure from business, the use commonly used methods of data analysis, case by breaking the novice help book. The book contains a large number of analytical methods, such as pest, swot, Boston's five forces, hypothesis testing, matrix analysis, correspondence analysis, and so, of course, which is the case with spss operation, we learned we have to take python python people realize again La! ! !
Followed by "python data analysis and data Operation" , which is a departure from the combat, describes how data analysis, data Operation of books with python, which contains a lot of tips, highly readable. Of course, the content of which involves a lot of machine learning, and therefore may take it as a learning book entry of the machine. Because the text is so small after reading this book, I began to machine learning journey.
Getting Machine Learning devaluation Andrew Ng's "open class at Stanford University: Machine Learning course" , when I was browsing in the NetEase cloud classroom, you can meet Dr. Wong finishing Chinese study notes together .
Followed by the classic "statistical learning methods" , currently a second edition, add the contents of unsupervised learning on the basis of the first edition. Often hear this dialogue:
A: "statistical learning methods" have any good? Why do so many people recommend?
B: What do you talk about this book, what is wrong? Feel bad, then read it several times it will feel good for him.
Finally, I reiterate, scikit-Learn official documents , both theoretical explanations, but also the use of methods, there are specific examples to help you quick start machine learning.
These are the current recommendations, if more books are interested can go take a look and see if can quickly analyze data to help get you started! Small text compiled e-books section, below concern the data of small text brigade , background replies learning books , you can get!
