What is the best training for Shenzhen data analyst? -Data analyst Skills Requirements

1. Basic Skill Requirements

Database knowledge (SQL must be familiar with at least), basic statistical analysis knowledge, Excel to be quite familiar with the SPSS or SAS have a certain understanding of the site-related business may also require to master GA and other web analysis tools, of course PPT is also necessary.
2 , data mining engineer
More is through the massive data mining, to find the existence of data patterns, or rules, so that through data mining to solve specific problems. Data mining is more about a specific problem, is to solve the specific problem-oriented. For example: Cluster analysis, through the members of various demographic, behavioral data analysis, to classify members, to different types of members to establish a corresponding profiling, so as to better understand the members, know the company member is exactly how? High, medium and low value of the membership composition, can not only later the operation of various members to provide guidance, improve the efficiency of activities, can guide the company's marketing, such as advertising strategy. and the development of various strategies for the company.
3. Key Skills Requirements

The database must be proficient. Most of the time, the data preprocessing of your model may be done in the database, and the database skills you use are higher. Must be mature data mining tools, data mining algorithms, such as Spss/celementine, Sas/em, of course, if you will 一、二款 open source software, and will write some program code that is the best, big companies like to use open source software, such as: R, WEKA.

4. Data Modeler

There is an essential difference between this position and the data mining engineer. Data modeler, more in favor of medium and small data volume, and its use is more statistical method, and data mining such as: Decision Tree, neural network, SVM and so on here is not involved.
Of course, the two have a common point, for very specific problems, are to solve a specific problem, such as: marketing response rate, you may be the history of mailbox, SMS response, to build models to predict, so as to improve the response rate of mail, or reduce the user's "junk" mailbox, improve the user experience. So from the mastery of the skills, the two are very different, the data Modeler in fact rarely mentions the word algorithm, more say what model to use, have feelings? However, from the practical perspective, these two models more and more do not have a clear division of labor, in general, will be two positions of people will learn each other's knowledge, so the two positions have a merger trend, but in the next few years, I think the company should be hiring should still have a difference.
New entrants into the data industry, can choose the corresponding position according to their own background, learning data, statistics, friends more can be biased to the modeler, and the computer especially write programming and classmates, can go data mining engineer, perhaps better adaptability, but this is not absolute.

What is the best training for Shenzhen data analyst? -Data analyst Skills Requirements