Chapter 11: The birth of the first product

奖励 A bonus column has been added to the main interface of the system. After Liu Fan opened the reward, there were two options in a row.
"Big data processing in-depth use": click to use directly into the brain.
智力 "Intellectual Development Manual": It cannot be substantiated, and you can click to check it at any time.
The reward is two things. It is simple and clear. Liu Fan is not entangled. He clicked on "Big Data Processing In-depth Application" first.
After touching your finger, the sense of form of the system has come again. I saw a book turning quickly on the screen of the mobile phone, and then gradually slowed down until it disappeared.
Almost when the book disappeared on the mobile phone, Liu Fan's brain felt a familiar comfort. A feeling of trickle flowing into the sea appeared in Liu Fan's brain, and it was not long before this wonderful feeling Disappeared completely, and then Liu Fan found a pile of knowledge in his brain.
This knowledge and Liu Fan's familiar ninety-nine multiplication table generally exist in the brain, very harmonious, there is no sense of abruptness, as if it is part of his learning from small to large, harmonious to make psychological preparation for the strength of the system Liu Fan couldn't help but be shocked again.
I digested the shocked emotion for a while, and Liu Fan quickly got a better understanding of the system rewards. The "Big Data Processing In-depth Use" provided by the system is a bit like a system architecture, but it is very detailed and very popular.
By analogy, it is like a home improvement design plan with full-process operation details. With it, you know what the electrician will do, the mason what the woodworker will do, how to integrate it in the end, and the overall effect. At the same time, , Materials, tools are ready for you, and all you need to do is to find masters such as electricians, carpenters, and masons. In the end, the quality and effect of the decoration of the house will be determined by the capabilities of the masters you are looking for Decide.
In the big data processing, the electrical carpenters and so on correspond to the knowledge of "discrete mathematics" and "probability". In other words, Liu Fan already knows how to design a big data processing tool, but the processing power of the tool he designed depends on his knowledge of related disciplines such as Discrete Mathematics and Probability. Reserve and understand.
This also means that the processing power of this tool is logically unlimited.
At this point, Liu Fan also finally understood why the system picked these three courses as tasks, because "Discrete Mathematics" and "Probability" are important basic subjects for developing related systems, and "Complex Function" can be optimized. Supplementary, of course, there are some very important subjects that are not included in the teaching plan this semester, such as Linear Algebra.
In addition, Liu Fan also noticed that in the aspect of programming, the system's reward is incidental, which is equivalent to the tools mentioned in the analogy. Only the basic mathematics is left as a space for Liu Fan to play. .
This has to remind Liu Fan of the sentence written in the system summary: I hope you can further understand the meaning of the task after receiving the task reward.
The system actually gives me a reward that is higher than my prayer needs, but at the same time, I need to continue to work hard. The harder I work, the more room for rewards to play. What does the system want to make myself understand?
Liu Fan still can't figure it out.
I do n’t think about it, I do n’t think about it at all, I plan to look at another reward first.
I opened the "Intellectual Development Manual" and came up with four words: Gengqin Naizhi.
Then jumped out a series of instructions. In short, Liu Fan can help his intellectual development by following the planned and continuous practice according to the manual, but this will be a long process and will not happen overnight.
Liu Fan knew what was going on, and there was no time delay. With system rewards, it is natural to start doing business. Now, first of all, we need to study several mainstream big data processing tools in the market, and then according to the principle of wooden barrels, It is necessary to make up for a few subjects such as "Linear Algebra" that have not been tested this year, and finally design your own tools.
With the basic idea, Liu Fan quickly entered the state.
In the summer vacation, my roommates all went back, just to give Liu Fan the environment for retreat and practice, there is no need to spend money and go out to rent a house or something.
After a short time, Bai Ku passed the gap.
三 At 3:30 in the morning on August 10, Liu Fanpeng walked unkemptly back and forth in the bedroom, looking excited for a long time.
For more than a month, Liu Fan almost stayed out of the house, but he could feel his rapid growth. On the one hand, it is an in-depth study of basic mathematics, and on the other hand, it is about the long-term plan for the commercial use of big data processing tools.
I chose big data processing as the entry point at first, and it was only a direction. The specific content was also planned in more than one month.
凡 The big data processing tools developed by Liu Fan naturally cannot just complete common functions such as automatic sorting or abnormal reminders. More specifically, it is a comprehensive big data service platform for enterprises, which covers internal management and customer management.
So Liu Fan's final result is essentially a framework that needs to enter the personalized needs of the enterprise to provide personalized big data processing services.
Liu Fan has also considered the possibility of making a universal tool. After all, this will avoid a large implementation cost in the future, but this kind of thing is not realistic when you think about it carefully. The situation of each enterprise is different. The needs are also different. To be universal, it is equivalent to meeting the needs of all enterprises in the world. I can't do it now.
Although Liu Fan did not have a technological breakthrough on general issues, the big data processing tools developed by Liu Fan have achieved huge breakthroughs in two other aspects.
The first one is the efficiency problem. Although Liu Fan still uses the random forest law as the basis for the algorithm, this is no way. With his current ability, he cannot change the current deep learning algorithm environment. After all, I did not consider the algorithm as the starting point for my own development when considering security issues, so I still have to use the current algorithm as the basis. But although Liu Fan has no ability to change, he has the ability to optimize.
凡 Liu Fan compared OpenRefine through a hundred experiments. The average operating efficiency of his system can reach more than 3 times. Again, this is a qualitative leap for enterprises with huge data. Take Penguin, for example, he generates hundreds of terabytes of data on WeChat and buckle every day.
What's more, the amount of data is increasing every day, because the speed of data processing can't keep up, more and more data can only be queued there, and many companies' decisions will have the problem of timeliness.
The second breakthrough is the diversity of data mining. The most important thing for the development of big data technology for an enterprise is his data analysis and data mining capabilities.
As one of the most classic cases in the big data industry: beer has the highest sales volume when it is placed next to urine, which is the substantial value that big data analysis brings to enterprises.
Enterprises want big data processing tools to tell themselves what can I do to make more money.
But on this issue, the current big data mining capability of the entire market is still in a relatively basic stage, because for data that has rules, big data is well analyzed, but more companies are going to be ruleless. It is difficult to analyze a result from the data. Because there are too many variables involved ~ EbookFREE.me ~ And through the improvement of system knowledge, Liu Fan thought of adding a data association inverse algorithm to the algorithm to let the messy data establish a functional relationship between themselves The function is mined twice to derive the effective function, and finally the physical function is analyzed.
This makes a lot of information that is judged useless in the current algorithm environment become valuable. Although this algorithm of Liu Fan cannot guarantee that the cluttered data can directly give the company an answer on how to make money, but it can allow the company to have a more comprehensive grasp of product conditions, market conditions, and company conditions. More traditional tools ignore The details will be presented with a high probability.
Nowadays, the homogeneity of various industries is becoming more and more serious. It is conceivable how important the utilization of information and the control of details are.
To say that the shortcoming of this tool is that the larger the amount of data, the more obvious its advantages. If the amount of data is too small, it may not show any advantages. This affects the problem of establishing target customers later.
多 After more than a month of study, R & D, comparison, and improvement, Liu Fan feels that this big data processing is now available as the first product.
Firstly, in terms of performance, Liu Fan thought that he had enough competitiveness. Secondly, for this system, Liu Fan had to rent a server for experiments and spend money on data. In just one month, he worked for three years to work. 20,000 yuan is almost spent.
I have to make money quickly.
As for the name of this tool, Liu Fan also thought about it. At that time, he picked it up because he looked like Mulong, so Liu Fan plans to use Mulong as the name regardless of technology, chip, or company.
This big data processing tool will also be the first product of Mulong Technology: Mulong Big Data Processing Tool 1.0.