Well, it's my first time to write an article in class, which is probably the first post-reading feeling I wrote after graduating from junior high school. There are two reasons for supporting me to write articles here: First, I have not consciously combed my thoughts into articles since I went to college, which greatly increased the inertia of thinking and failed to cultivate the habit of systematic thinking, which is a pity; Secondly, I adopted Markdown's writing typesetting method. Only recently did I know the advantages of this writing, and I want to take this opportunity to learn.

So try to get into the habit of recording your thoughts here in the future.

Anyway, if you are a liberal arts student, I suggest you try to read this book when you are free. It belongs not only to engineers, but also to something worth pondering and exploring in this world.

When I was in college, I was interested in stock speculation. Every day, I am thinking about how to get rich overnight in this near-zero-sum game market, and I have checked a lot of tricks of unknown people. As a result, as you guessed, it was useless. It was not until I read several books on how American hedge funds operate that I suddenly realized: Fuck!

The best stock market participants in the world no longer rely solely on the information intake and judgment of the human brain to make investment decisions. They rely on advanced computers and trading models to make profits. If you also set foot in this field, you should have heard of the famous Renaissance technology company, and the core of this advanced trading method is: mathematics. Or closer to the times, called big data.

With his profound mathematical skills and long-term project development experience in first-tier Internet companies such as Google and Tencent, Mr. Wu Jun described the mathematical principles of Internet functions such as search, translation, navigation, speech recognition, web crawler, web ranking and anti-cheating in extremely simple language, and conducted an approachable science popularization for those jobs that only belong to engineers and scientists in the eyes of ordinary people. Often, a mathematical equation reveals many terms that we feel tall in our daily life, such as "artificial".

Personally, I looked at it with an exploratory interest at first, and I didn't want to accidentally read something related to automated trading.

Like the maximum entropy model, Markov chain, Bayesian network and artificial neural network have made great contributions to today's automated transactions. These mathematical models and ideas have gradually entered the American investment market since the 1990s, and their achievements are beyond the reach of companies like Berkshire Hathaway.

After all, the information that a human brain can process is too limited. Even the wisdom of several traditional fund managers can't compare with the energy of the whole market. However, with the help of mathematics and today's powerful data collection and calculation capabilities, it is possible for us to quantify countless influencing factors and make accurate judgments.

Another question that makes me think is how to train the parameters of mathematical model and use data to promote function iteration when there is no data or only a small amount of data. Because of a problem in our recent work, the development of a product in our hand is almost completed, but the core function of this product needs a batch of data. If we have enough user traffic, we can iterate the core functions of the product according to the user's data feedback. However, due to the design problems of our products, if we completely rely on User Contributed Content (UGC), the user experience will be seriously affected, which is the cold start problem. As the first batch of users (or hired people), it is too expensive for us to do UGC in the corresponding scene. This forced me to think about how to use some algorithms to improve the product as soon as possible with less data. The beauty of mathematics also gives some cases of "out of nothing", such as PageRank algorithm that helped Google become famous in World War I.

Finally, two simple concepts are described. What is encoding and decoding? During my first job, as a liberal arts student, I had a headache about these two concepts and related problems, and I found the answer in this book. For a popular example, we express our ideas in words, which is coding. A person who listens to us absorbs what we say and understands it in his brain. This is decoding. Maybe everyone thinks this is a very common process. Isn't this a natural thing? Think about it. Why can things in the brain be expressed in words or written in words? The information stored in words written on paper is completely different from that stored in the brain. This set of conversion rules is actually coding and decoding, while English and Chinese are two different sets of coding and decoding rules. Similarly, when we make a phone call, the acoustic information we send needs to be converted into electrical signals, transmitted to each other by radio, and then converted into auditory signals that people can understand. This is also a process of encoding and decoding. The encoding and decoding of all forms of information is also a mathematical work in essence. More directly, when we input information through the computer, the most common way is typing. However, when these messages are encoded and handed over to the computer, they are all stored and transmitted in binary.

Generally speaking, mathematics plays a very, very important role in our world, and all our daily work is inseparable from mathematics. This is a great experience from my recent work and study. Learning to cultivate my interest in mathematics and accumulate mathematical theory and knowledge is probably something I will do seriously in my life.