Data analysts have gradually become a popular position. In this era when traffic is king, the importance of data is also increasing. Many companies are aware of this and have begun to recruit relevant data-oriented talents.
The basic career development path of data analysts is: junior data analyst-intermediate data analyst-senior data analyst.
The data analysts in these three different stages are progressive in terms of job content, core skills, and salary. This not only shows that data analysts need to maintain the ability to keep learning, but also shows that the position of data analysts is a rising and very promising position. job development prospects.
So what skills do you need to have and what business do you need to handle at each stage? Let's take a look together.
Junior Data Analyst
There are three most common tasks in primary data analysis: data extraction, report development, and writing analysis reports.
Data extraction will be your main job and the basis of all analysis.
It is no exaggeration to say that 80% of the energy in most analysis projects is in the stage of data acquisition and processing.
Trying every means to extract data from the database seems not complicated, but it is a test for your SQL skills.
Sometimes an over function can save hundreds of lines of code, and a piece of SQL optimization can increase the efficiency by more than ten times. So do a good job of data extraction in a down-to-earth manner, and don't think this is meaningless.
After you can extract data efficiently and accurately, you can start learning report development: solidify common data extraction to form tables or visual charts.
At this time, you will find that the previous knowledge of Excel and database design will come into play.
It is necessary to further study the basic knowledge of BI, and understand what star schema, data warehouse, cube, etc. are. Recommend "Data Science and Big Data Analysis".
Excel is also a visualization tool, but it can only be viewed on a single computer, so more often you will learn some report development tools or visual analysis tools, such as PowerBI, Tableau, etc. You can find video tutorials of these tools on the Internet, just download the trial version and practice by yourself.
Writing an analysis report is to combine many reports into a PPT that can be read by people.
At this time, the test is your PPT skills. In addition to concise and beautiful PPT, a good analysis report is more important to grasp the theme, clear analysis ideas and valuable conclusions. If you can still vividly tell the content of the PPT, it will be even more perfect.
If you have done the above three aspects well, congratulations, you should have become a data analyst with a little success.
Intermediate Data Analyst
On the basis of primary requirements, it is required to master theoretical knowledge such as multivariate statistics, time series, and data mining, master advanced data analysis methods and data mining algorithms, and be able to use at least one professional analysis software such as SPSS Modeller, SAS, Python, and R proficiently.
Familiar with the application of SQL to access enterprise databases, combined with business, can extract relevant information from massive data, conduct modeling analysis from different dimensions, and form data analysis reports with strict logic that can reflect the overall data mining process.
Database technology (compulsory): Use SQL to access enterprise databases, combined with business, can extract relevant information from massive data, conduct modeling analysis from different dimensions, and form data analysis reports with strict logic that can reflect the overall data mining process; simple understanding of relational The relationship between database and non-relational database, database addition, deletion, modification and query, advanced query, advanced application
Practical big data mining algorithm, (Apriori algorithm, Tanagra tool, decision tree): explain data mining technology in depth. The so-called "deep" refers to starting from the principles and classic algorithms of data mining. One is to understand algorithms and know which method should be applied in which scenarios; the other is to learn the classic ideas of algorithms, which can be applied to other practical projects; the third is to understand algorithms, so that data mining algorithms can Apply to your project development. The so-called "simplified" refers to the implementation of the application of data mining algorithms into practical applications. The course will explain the application of algorithms through three different aspects: one is data mining realized by tools such as Microsoft SQL Server and Excel; the other is data mining of famous open source algorithms, such as Weka and other open source tools; the third is using Java, C# language and two languages are demonstrated to complete the realization of the data mining algorithm.
SPSS Modeler data mining: connect the ideas, methods, and parameters involved in mining technology with the basis of statistics, and understand the functions including dimensions, data, analysis, and data flow, the actual meaning of parameters, and application methods such as selection and combination .
Python web crawler technology: master the application of Python crawler basic library; master the use of Python crawler tool; master the use of Scrapy project construction; master Scrapy streaming development; master the use of Scrapy expansion; master the use of Scrapy to interact with Mysql.
In-depth machine learning extension (Python language, algorithm, Numpy library, MatplotLib): two types of machine learning methods: supervised learning and unsupervised learning, of which supervised learning is divided into classification and prediction of numerical data. These algorithms are basic algorithms. Realize the underlying algorithm of data mining by learning Python in depth.
Machine learning for artificial intelligence (extended): Understand linear regression, master the application of decision trees, proficiently use SVM support vector machines, proficiently use clustering + Bayesian, master EM-HMM-LDA-ML.
Article source: The network copyright belongs to the original author
The above content is not for commercial purposes, if it involves intellectual property issues, please contact the editor, we will deal with it immediately