Mathematical knowledge is the basic knowledge of data analysts.

For junior data analysts, it is enough to know some basic contents related to descriptive statistics and have certain formula calculation ability, and it is better to know commonly used statistical model algorithms.

For senior data analysts, knowledge of statistical models is an essential ability, and linear algebra (mainly knowledge of matrix calculation) is best understood.

For data mining engineers, in addition to statistics, various algorithms also need to be skillfully used, and the requirements for mathematics are the highest.

Beginners need to have two basic analytical tools to learn big data.

For junior data analysts, you need to be able to play Excel and skillfully use pivot tables and formulas, VBA is better. In addition, you must learn a statistical analysis tool, and SPSS is better as an introduction.

For senior data analysts, using analytical tools is the core competence, VBA is the basic necessity, SPSS/SAS/R should be proficient in using at least one of them, and other analytical tools (such as Matlab) depend on the situation.

For data mining engineers ... well, Excel is enough, and the main job is to write code.

Beginners need to have three basic programming languages to learn big data.

For junior data analysts, you can write SQL queries, Hadoop and Hive queries if necessary, basically.

For senior data analysts, besides SQL, they also need to learn Python, which is used to acquire and process data, and get twice the result with half the effort. Of course, other programming languages are also possible.

For data mining engineers, Hadoop must be familiar with at least one of Python/Java/C++, and Shell must be able to use ... In short, programming language is definitely the core competence of data mining engineers.

Beginners of big data need to have four basic business understandings.

It is no exaggeration to say that business understanding is the basis of all the work of data analysts. The data acquisition scheme, the selection of indicators and even the insight into the final conclusion all depend on the data analyst's understanding of the business itself.

For junior data analysts, the main job is to extract data, make some simple charts, and a small number of insight conclusions, and have a basic understanding of the business.

For senior data analysts, they need to have a deeper understanding of the business, and can extract effective opinions according to the data, which is helpful to the actual business.

For data mining engineers, it is enough to have a basic understanding of the business, and the focus still needs to be on exerting their technical ability.

Beginners need to have five basic logical thinking when learning big data.

This ability was mentioned less in my previous article, so I took it out alone this time.

For junior data analysts, logical thinking is mainly reflected in that every step in the process of data analysis is purposeful, knowing what means they need to use and what goals they need to achieve.

For senior data analysts, logical thinking is mainly reflected in building a complete and effective analysis framework, understanding the relationship between analysis objects, and knowing the cause and effect of each indicator change, which will bring impact to the business.

For data mining engineers, logical thinking is not only reflected in business-related analysis work, but also includes algorithm logic, program logic and so on. Therefore, the requirements for logical thinking are also the highest.

Beginners need to have six basic data visualizations to learn big data.

The requirement of data visualization is high, but it actually covers a wide range. Putting data charts in a PPT is also regarded as data visualization, so I think this is a universally needed ability.

For junior data analysts, Excel and PPT can be used to make basic charts and reports, which can clearly display the data and achieve the goal.

For senior data analysts, it is necessary to explore better data visualization methods and use more effective data visualization tools to make simple or complex data visualization content according to actual needs, but it is suitable for the audience to watch.

For data mining engineers, it is necessary to know some data visualization tools and make some complex visualization charts as needed, but usually there is no need to consider too many beautification problems.