First of all, let me declare that the following introductions in this article are my own opinions and experiences; they are all spoken in plain language , and will not introduce any normative concepts.
1. What is mathematical modeling?
To put it bluntly: let us use mathematics to understand the world (pure nonsense).
In fact, to put it bluntly: it is to use mathematics to explain and solve various problems in life (such as stock forecasting, fire alarm statistics, etc.). (In fact, you don't need to give a definition to mathematical modeling, just simply understand it as using mathematical methods to solve problems)
2. Why do you want to participate in mathematical modeling?
Perhaps for some great gods with high sentiments, it is to promote the development of the motherland and the progress of society.
But for an ordinary person like me, bonus points! Extra points! Extra points! Keep a graduate student and go brag about B with your friends!
For students who want to maintain their research, the mathematical modeling competition is a must, and it will really add a lot of points, which is more effective than doing research and development or applying for patents, so I suggest that everyone must pay attention to this competition. How many points to add, each school has different indicators.
3. Mathematical Modeling Competition
There are many competitions in mathematical modeling. Here are a few well-known competitions.
1. "Certification Cup" Mathematics China Mathematical Modeling International Competition
Sponsor: Inner Mongolia Autonomous Region Mathematical Society and Global Mathematical Modeling Ability Certification Center
Deadline for registration: 00.00 on December 3, 2021
Competition time: 8:00 am on December 3, 2021 - 8:00 am on December 07, 2021
2. American Collegiate Mathematical Contest in Modeling
Sponsor: American Federation of Mathematics and Its Applications
Registration deadline: before 4:00 am on February 18, 2022
Competition time: 6:00 am on February 18, 2022 - 9:00 am on February 22, 2022
3. MathorCup University Mathematical Modeling Challenge
Sponsor: China Research Association of Optimizing Laws, Overall Planning and Economic Mathematics
Registration period: 0:00, December 30, 2021 to 12:00, April 13, 2021
Competition time: 8:00, April 14, 2022 to 9:00, April 18, 2022
4. "East China Cup" Mathematical Modeling Invitational Competition for College Students
Sponsor: School of Mathematical Sciences, Fudan University
Registration time: 9:00, April 18, 2021—18:00, April 30, 2021
Competition time: April 30, 2021 - May 4, 2021
5. "Huazhong Cup" Mathematical Modeling Challenge for College Students
Sponsor: Hubei Society of Industrial and Applied Mathematics
Registration period: March 15-April 29, 2022
Competition time: 20:00, April 29, 2022-20:00, May 2, 2022
6. May 1st Mathematical Contest in Modeling
Sponsors: China University of Mining and Technology, Jiangsu Society of Industrial and Applied Mathematics, Xuzhou Society of Industrial and Applied Mathematics
Registration time: 00:00, April 1, 2022—24:00, April 30, 2022
Competition time: 9:00 am, May 1, 2022—9:00 am, May 4, 2022
7. National Undergraduate Electrical Engineering Mathematical Contest in Modeling
Sponsor: Electrical Mathematics Committee of the Chinese Society for Electrical Engineering
Registration period: April 11, 2022 - May 25, 2022
Competition time: 8:00, May 27, 2022-8:00, May 30, 2022
8. "Shenzhen Cup" Mathematical Modeling Challenge
Sponsors: Shenzhen Association for Science and Technology, China Society for Industrial and Applied Mathematics
Competition time: July 15-September 10, 2022 (refer to 21 years)
9. "Higher Education Society Cup" National Undergraduate Mathematical Contest in Modeling
Sponsor: Chinese Society of Industrial and Applied Mathematics
Registration time: The deadline is 20:00 on September 2, 2022
Competition time: 18:00, September 15, 2022 to 20:00, September 18, 2022
For more information, visit the competition official website.
4. How to prepare for the Mathematical Modeling Competition
How to learn mathematical modeling has become the main concern of everyone. When I was studying, I also read a few recommendations from other seniors, and learned according to their ideas, for example: I made a lot of book lists for you to read, but believe me, you can’t read it at all. A few pages, even thirty or forty pages, I admire you.
Here's how I think about learning:
1. Go to station b to watch the video
First of all, it is necessary to figure out what the various models are for, what are their characteristics, and in what scenarios are they most suitable for use. (For example, the topic is an analysis and evaluation question: is it suitable to use TOPSIS, principal component analysis, or cluster analysis.), at least you must know what these models do, so that when you see the topic, you will "Oh , This question is suitable for neural network machine learning prediction; Oh, that question is suitable for principal component analysis", and then go to find the information of the selected model below, which reduces a lot of workload. Recommend "Qingfeng Mathematical Modeling" and "Brother teaches you how to model".
2. Follow the examples in the video to practice
When explaining these videos, there will be some sample questions and after-school homework, and then demonstrate how to do it (such as Matlab, spss, stata). Don't just watch, you must follow along, and you will find many problems in the process of doing it.
3. Directly criticize the excellent papers of the national competition! ! ! ! ! ! ! ! ! ! (Believe me)
After watching the video and understanding various models, I went straight to the paper. It must have been painful at the beginning (ah, an article has twenty or thirty pages, how can I have the patience to read it; ah, what is this writing, completely can't read).
I remember when I was reading the thesis, I thought, "This is meow, is it written by a human being?" I lost my patience after reading it, and my mind was distracted, and then I patted my head and continued to bite the bullet. But when you read three or four articles, you will discover the essence of them, see how others analyze the problems, how to choose models, and how the pictures they make are so beautiful, and the articles are so organized.
Then just follow along, learn how others analyze problems, learn how others typesetting, learn how others make pictures look good, isn't it all your own after learning.
4. Participate in a large number of competitions! ! ! ! ! ! ! ! ! (military training)
Don't think about what I'm doing if I don't know anything about participating in the competition. If I have that time, I might as well watch two more videos and read two more articles. I tell you, such thinking is a big mistake,
It doesn’t matter if you don’t know how to do it, learn according to the topic, and even if it’s editing, you have to compile the thesis. You will feel it after participating in a few competitions. At the beginning, you must have a good attitude and don’t blindly pursue awards. We are just training soldiers, etc. If you improve your knowledge, you can basically win all the awards in one year.
5. Team insights
As the saying goes, "Three cobblers are the best Zhuge Liang", the modeling competition is also a team of three people, generally speaking: one person is in charge of programming, one is in charge of modeling, and one is in charge of the thesis.
For big data questions, my opinion is that two people are responsible for programming and one is responsible for papers. One of the two programmers is mainly responsible for data processing, and the other is responsible for the writing of model algorithms. Why do you say that, after publishing the topic, three people must discuss it together, express their understanding of the topic, and then find the best method. In this process, the algorithm has basically been determined, and the workload of subsequent modelers is basically not It's big, but the workload of programmers is huge, and they also need to process data and run algorithms to get results, so I suggest that teams preparing to do big data problems can consider the following team formation.
For big data topics, the best team insights:
(1) The main algorithm programming team
This team member needs to have a deeper understanding of the algorithm and have his own knowledge and understanding of various models.
In the process of discussing the topic, collect everyone's understanding, judge the feasibility of various ideas, and find information to write the program of the model after the model is determined.
(2) Master data programming team members
For big data topics, there will be a huge amount of data to be processed (I remember this year’s May 1st Mathematical Modeling Contest, which has more than 100,000 rows of data), and there are still some data dimensions that are different (such as input with "day" as the node , the output takes "hour" as the node) and other issues. It can be done by writing pseudocode in excel, or using Matlab or python.
(3) Members of the main thesis team
The team members of the main thesis must have good writing skills, as well as beautify the aesthetics of typesetting. Do a good job of data visualization (make some good-looking graphs, and store the data in the "three-line table" format); read more excellent papers, and summarize the writing experience of others.
Of course, these three people must have a certain understanding of various models, otherwise the analysis is wrong, the correct algorithm cannot be determined, and what program to program. Therefore, it is not only necessary for one person to know the algorithm, but everyone must Knowing algorithms, including paper writers, if you don’t understand algorithms, you can’t write anything at all. If you don’t have your own insights, you can’t just go to the Internet to find articles ctrl+c, ctrl+v, and you may not even be able to pass the plagiarism check of the paper.
There is also a more important point, everyone, don't keep thinking about "finding a big bull to hug your thigh", it is best to find someone who can get along with you in the team, and you really want to improve yourself through this aspect. It’s not that there is no problem if you know more. You can learn together and improve slowly together. So you must be cautious when looking for teammates. I have seen too many teams that fell apart before the end of the game.
I will continuously update the knowledge about mathematical modeling, as well as the "tips" that everyone is most concerned about, share my personal experience, and help everyone win awards better.