But Python also has many advantages. It has a simple structure and does not need a lot of syntax details to learn. It allows many non computer professionals to write scripting programs. For example, financial practitioners used to write transaction scripts, engineers used to write drawing processing scripts, and data scientists used to write image analysis scripts, which gave this language a better mass base, so many good third-party libraries were slowly developed by people. For example, pandas and matpolotlib are most commonly used by data scientists, yfinance and tushare are most commonly used by financial practitioners, numpy is most commonly used by engineers, and so on.
But most of the scientists engaged in artificial intelligence are not professional experts in software. Why not use it when they find that a computer language is simple enough and there are many good third-party libraries that are free and do not need to write code from scratch?
Python has many third-party libraries. How many are there? 340000, the largest of all computer languages. The so-called third-party library refers to the small code that others have developed. For example, I want to write a software like wechat. If I lack experience, I often find a way to write it myself and constantly organize syntax, variables, functions and conditional statements. Sometimes, a lot of heart misses a semicolon and writes an extra colon, and the program can’t run. But at this time, I found a third-party library with complete functions and mature code. I don’t need to think about it. I can directly reference it, and then change the code myself, and the program will run successfully.
If you are an algorithm master of artificial intelligence, do you want to spend more time on improving algorithms, programming and building frameworks? At this time, python has become the first choice for artificial intelligence development.
Now, many training classes use Python to rub the heat of artificial intelligence and publicize “Python zero foundation introduction, Xiaobai can learn Artificial Intelligence in three days”, “artificial intelligence training camp, one line of code teaches you to open artificial intelligence mode”, etc. This is only a false publicity to attract students, which is a serious misleading to the public.
Learning Python is not equal to artificial intelligence, that is, using C / C + + computer languages can replace most of the work of python, but if you want to improve yourself by learning python, this computer language is still very good. For those of us who do finance or auditing, even if we can write code, we won’t have too high requirements for code. Just run. Don’t learn all the profound knowledge such as computer theory and code analysis. We are finance, we are audit, and our work is not computer engineers. We can master a certain level of operation, I had no experience before. I learned a lot of useless knowledge and wasted time.
Since this article is not a special explanation of the usage of Python, if interested friends, you can look for my historical article on the official account, search for “frozen apples without worms”. There are many examples of Python tutorials and application writing, and some cases of artificial intelligence are also introduced. Address:github.com/Gandedong/audit-python
There are a lot of artificial intelligence development in the market written in Python. Python is the best way to understand the ideas. If you are not engaged in artificial intelligence, python, as a computer language with rich third-party libraries, can also help you in other work. For example, python has pandas library, which can help you do pivoting; There is an xlwings library that can help you consolidate reports in batches; There is a Matplotlib library that can help you make charts; There is a requests library that can help you download reports; Pypdf2 library can help you make PDF files; There is a MySQL Python library that can help you connect to the SQL database; There is a pyautogui library, which can help you automatically record vouchers; The above can be applied in financial work, or improper accounting, or go to financial companies to engage in relevant work.
Python can also be used to build asset pricing models, risk management, and quantitative transaction management. Input Python + Finance on the recruitment website of Beijing, Shanghai, Guangzhou and Shenzhen, involving 2200 positions.
In addition, the most important thing to learn a computer language is to practice it. The best way is to follow others to knock the code one by one. Just like learning to drive, it may be slow and difficult at the beginning, but over time, it will drive fast. After all, compared with the knowledge of linear algebra, statistical learning, neural network, modeling and recognition of artificial intelligence, programming is the easiest.
This work adoptsCC agreement, reprint must indicate the author and the link to this article