Technical thinking-don't think about big data from a technical point of view

After defining their specific modules in the direction of big data, only to strengthen their own this aspect of the breakthrough and technology accumulation. I choose to be data analysis and visualization in two directions. Industry is the traditional industry, why is this? The following is a layer of analysis from a business perspective.

All along, the technology that is too advanced is not! It is just your usual accumulation and details of the control. and data analysis, now all walks of life need, small to year-end financial summary report. Large to the company's CRM system. But the traditional industry is in a wait-and-see attitude towards new technologies. The traditional profit model has made the traditional industry afraid of the industry that the Internet is changing at this moment. So, data analysis, for big data in the traditional industry, is a good point of entry. And how to let the other side perceive the benefit, this, is also very important! Remember my data mining teacher Zhang Lei (Data mining field Daniel) told us that in the traditional industry, if your data mining does not allow customers and business units to see benefits in six months, then you will be a very failure of the data Analyst!

The Chinese are good at summarizing the process and see around corners. Speculate on what is going to happen, based on a lot of observations. This is also a data analysis, and really associated with big data, may be the user can directly interface with the data. And this some data pre-finishing, cleaning, analysis ..... are basically in some Excel, or some txt,word. At this point, your business analysis logic is very important, that is, you have to have a good understanding of the user's data business information. When later, when the user or business unit discovers the value of your data analysis, this time, you don't have to say, they will give you a lot of data to let you analyze.

When you have more data, you can consider the relevant technology of the database, how to access your original Excel, or TXT document. At this point, you can try a structured database. At this point, the technical input is estimated to be the database purchase, but its investment will be much smaller than the value you create to the company's business. How data is fetched, how it is read into the database, is what you have to consider and important. Also tell the business section where their critical data is accessed.

When you analyze to a certain extent on a single machine, the data becomes more and more, and you start to think about the data of multiple dimensions. Don't speculate on your target customers by just taking a single dimension of data. In this way, it will have effect, but the effect is very low, think of those who recommend ads to you system it! is not quite a failure. And the film industry's "Guess you like" this option is more valuable to you. Well, when you start to bring value to your customers from multiple dimensions, you'll start to get some idea of the algorithm, after all, this piece is directly determining the success rate of your results is high or low.

Data more and more, a storage is not stored at all how to do! The next step is to build a distributed architecture, remote disaster recovery, mutual preparation. Server, routing, some of these technical concepts are about to start, because next, you'd better learn the system architecture. For example, on the network, load balancing, TCP unilateral acceleration, TCP proxy. The server in the wave-yong protection, double Live center. Storage aspects of the disk array, SSD storage. High-performance CPU computing .... In this step, you will start to strengthen your technical skills.

Some people say, now all began to cloud, these I do not have to care, all managed not to be able. I'm only in charge of running my business, all right? But really start running data on the cloud, will you not be able to recruit a person to manage it? The development of technology, will only eliminate the low-end network and IT operations, and began to be high-end operational needs more and more.

All right, let's write this much today! It is also a small sort of thinking about oneself. Next time, let's write this topic down.

This article is from the "Data Mining and Visualization" blog, reproduced please contact the author!

Technical thinking-don't think about big data from a technical point of view