Shulou(Shulou.com)06/01 Report--

How to start big data analysis, many novices are not very clear about this, in order to help you solve this problem, the following editor will explain for you in detail, people with this need can come to learn, I hope you can gain something.

For the rookie of data analysis, I have the following suggestions:

First, the technology needs to be excellent.

Sql sentence, you have to learn to master, add, delete, change and query, and stored procedures you need to be proficient in everything.

The purpose of python,python learning is to make it more efficient, convenient and fast for you to deal with complex and large data sets, and there are a lot of machine learning or deep learning applications that you can start with.

Second, master or master some commonly used mathematical principles

Basic mathematical principles, statistical related principles, such as mean, variance, standard deviation, derivative, inverse function, Lagrange and so on, here you can go to csdn or Taobao to buy a related book is enough.

The principle of machine learning algorithm: regression algorithm, classification algorithm, clustering algorithm and so on, such as the simplest classification algorithm: decision tree, do you really master its mathematical principles? Do you know what information entropy and conditional entropy mean and how to calculate them? If you can only switch packets, then solving the simple problem ok involves a lot of complex situations, such as uneven samples, such as too few features, you can't start. In this section, I recommend you to visit the forum more often, which is very useful and can learn a lot.

Start to do some virtual cases, or participate in some projects of colleagues in the group, such as an in-game advertising chat to pull people to identify the project, its essence is to achieve text recognition through classification algorithms, this piece of knowledge may be: text processing, classification algorithm applications, etc., this piece can go to the github to take a look around, there may be some cases that may have been operated to give you inspiration.

Third, constantly strengthen your perception of products.

Experience the product with problems, mainly push back what's good and what's wrong with making the product like this, what problems the product is mainly to solve, and what his business goals are likely to be.

Often discuss and communicate with people in the circle, unlike games, there are many salons on the Internet, you can participate in this kind of salons, but do not be miscellaneous and extensive, but be specialized and skillful. Now there are a lot of salons just to do it, to move yourself, there is no need to participate.

Always pay attention to the competition, such as you are a social products company, then Wechat or Momo, every iteration of the version, you need to think about what problem he is trying to solve, and then combined with the specific content to evaluate whether his goal has been achieved.

Accumulate more data, think more and ask more questions. When we do specific data analysis, we sometimes need to respond quickly to the business, and sometimes it is difficult to find problem points through traditional analysis methods. Accumulation is very important at this time. 40% of product decisions in the game industry rely on data, 30% on user feedback, and 30% on experience.

What is accumulation? Accumulation is the relationship between product adjustment, activity design, and data in the past. Only on the basis of serious and rigorous treatment of every work, can this accumulation be done well, so that your products and data will become more and more acute. later, you can cross-border to become a product expert.

The positioning of ourselves is not only a data analyst, but also a half-product or operations leader. We have a lot of data analysts who are in a very passive state all the time. You can do whatever the business requires. Over time, you will become a person who mechanically meets the requirements to complete the requirements. There is no sense of existence. How to change this situation?

First: to do more accumulation, as mentioned above, this area does not need time.

Second: you need to take the initiative to treat yourself as half the head of product or operation, so that when you face daily data, you will take the initiative to improve your daily monitoring indicators, and you will stare at the anomalies and changes of the data every day. You will dig deeply according to these anomalies and changes, or when the data is not abnormal, analyze what to do according to experience at this stage and do it quickly. If you find a problem or potential risk, talk to the business about the solution quickly, so that you can maximize the value of the data analyst.

Is it helpful for you to read the above content? If you want to know more about the relevant knowledge or read more related articles, please follow the industry information channel, thank you for your support.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.