Link:/question/36214681/answer /66483598

Source: Zhihu.

The copyright belongs to the author. Please contact the author for authorization for commercial reprinting, and please indicate the source for non-commercial reprinting.

Let's start with business analysis. I followed many similar questions and answers on Zhihu and found a serious problem: most people's understanding of the business analysis industry is not accurate. Specifically, there are two kinds of business analysis:

1. Business analysis. It is a traditional business, and its main task is to analyze the whole operation process and business development of the company. For example, discover new business requirements and propose/improve solutions to some business problems. It may include organizational change, business process improvement, strategic planning, policy formulation and improvement. This field needs some data analysis, on the one hand, because companies generally rely on business expansion in the early stage of development, which will involve some similar data analysis such as sales performance. On the other hand, in the era of big data, many companies feel that they have to follow suit. But in general, it is still based on business analysis. That is, qualitative analysis is given priority to, supplemented by data analysis (generally simple). This is also mentioned by most interviewees. To grow into a professional in this field, MBA is generally chosen. MBA courses generally include some related courses, but they are very simple.

2. Business analysis. This is the real emerging discipline. Translation is called business analysis, and the content is much worse, so it is deliberately distinguished in English. The core of this industry is data analysis, which is an advanced technology, model and algorithm. Through in-depth analysis and mining of the data, we can study the company's past performance and search for potential business information in the industry market. The purpose is to gain insights that qualitative analysis and simple quantitative analysis can't. This field is gradually attached great importance by academia, industry and government. In academic circles, since new york University started to offer a master's degree in business analysis in 13, many universities around the world have added related majors, and the tuition fees are almost the same as MBA. In capitalist countries where money is paramount, high tuition fees are often the vane of high income and good career development. The industry, not to mention BAT, is opening such a department. Last year and this year, Alibaba hired some professors in this field from famous universities, and Baidu invited Daniel Andrew Ng from Stanford University. Government: As far as I know, the governments of Singapore and Australia directly sponsor companies that set up business analysis departments.

After talking about the basic concepts, let's get back to the point. After reading the problem description carefully, I believe that the subject should be interested in business analysis (if not, please ignore the following). Then talk about the related skills in a little depth. The topic is undergraduate statistics, which is very good! Ha ha! I am a doctor of statistics, and now I am doing a similar job in an investment management company in new york. My colleagues include some doctors in data analysis related fields such as computer, automatic control and signal processing. Generally speaking, business analysis is undoubtedly an interdisciplinary subject, including mathematics (statistics), computer and business (economics, marketing, game theory, etc. ). Statistical methods occupy a primary position in the discipline of data analysis (mainly regression models). It can be said that if all statistical methods are taken away, data science is basically fragmented. Then there is the computer (including machine learning, pattern recognition, image processing and other fields). Based on the requirements of some application levels, the computer field has put forward some novel ideas and models. It is worth mentioning that these things have attracted the attention of statisticians, who have solved the same problem by statistical methods. Combined with some traditional regression models, a new branch of statistics has emerged: statistical learning. Finally, business is the smallest, but some ideas are also worth learning, such as game theory. Some specialized masters of business analysis will talk about statistics and computer methods in combination with business applications.

Actual business analysis cases (business analysis, of course). Simply put: 1. Survival model is the most commonly used model in biostatistics. Studying the healing and death time of diseases can be used to test whether some drugs and treatments are effective. When it is applied to business, for example, companies advertise to specific groups of people, and they click and watch through different channels, then how long will it take them to decide to buy? Our definition: consumers don't buy =' alive', once they consume, they are' dead'. The next task is to study which advertising channel/combination is the most effective. 2. As for the cluster problem, the company put in many advertisements and made many promotions. Then a large number of consumers came to buy, which ones really saw the above publicity? If you don't know how to analyze which channel or combination of channels is the most effective. Cluster analysis is to solve this kind of problem. 3. Bayesian method, I wonder if you have ever known Bayesian analysis? It is a very popular direction in the field of statistics/machine learning. The main application is that artificial views can be integrated into the model when modeling. Combine qualitative and quantitative analysis results. This is very popular with people who do data analysis in the field of business analysis. There are many others, so I won't talk about them one by one.

Finally, talk about your recent situation. The undergraduate background is still a little weak. If you want to get in touch with the above kinds of jobs, you need to go further. And if you want to eat well, it is best not to be confined to the field of statistics. After all, learning an applied subject and learning to consider problems from the perspective of application is more beneficial to career development. If you want to enter a higher school, you can consider studying for a master's degree in data analysis/business analysis. However, if you think the tuition is too high, in fact, a master of statistics is also a good choice. If you teach yourself, there are also many good courses on coursera. As for R/Python/SAS, knowing one is enough. Data analysis level programming, you can know one, and you can get started in another week.

Finally, share a link: graduate programs in big data analysis and data science. It lists the postgraduate majors in data analysis offered by universities around the world, as well as various paid/free online courses.