Due to various reasons this year, for the first time, and it is said that it will be the last time, the US competition will be temporarily adjusted and a spring competition will be added. This is a great opportunity for everyone in desperate need of a modeling award. Whatever the reason, we may have some regrets in this year's competition. However, the spring split may be an opportunity to make up for regrets. First of all, I will bring you a brief analysis of the Z questions in the spring competition, so that you can choose better topics. And predict the difficulties that will be encountered in different competition questions in advance, so as to avoid lightning in advance.
Question Z detailed ideas eight pages word
Link: https://pan.baidu.com/s/1blv8pZK6X3d6s3L0ETd1Ow
Extraction code: sxjm
2023 ICM Issue Z: The Future of the Olympic Games
The topic setting of question Z directly reminded me of question B of the 2010 National Competition , when the topic was set as the quantitative evaluation of the Shanghai World Expo. The setting of the background and the method of asking questions are self-evident. Therefore, I will collect my 2010 national competition data for you to share later. Below, a brief analysis of the competition.
Innovative problem solvers are considering various options and strategies to address these issues due to the reduced number of bids for the Summer and Winter Olympics. For example, an idea for both the Summer Olympics and the Winter Olympics should have a fixed location. Another idea is to divide the Olympic Games into four (rather than two) groups and have four smaller Games (e.g. winter, spring, summer and autumn). Such a system may ease the burden of staging such a large event to some extent.
In order to look at the topic more intuitively, we divide the topic setting into three parts, namely the selection of indicators, question 2, and non-technical articles.
Question 1, the selection of indicators , and we need to conduct problem-solving analysis based on these two ideas, we need to analyze from the economy, land use, people's satisfaction (athletes and spectators), travel, future improvement opportunities, host city/country Prestige, an indicator for establishing hosting the Olympic Games.
Data collection is a reference to the problem. For these indicators, we first need to collect relevant data. I will collect some data to share later here. When you collect by yourself, you should pay attention to collecting data and only need to grasp the general direction. We first collect the data of the topic and give various general directions. According to the collected data, the selection of the index of question 1 is completed. Remember, you can’t select the indicators first and then find the data, so it is easy to fall into the embarrassing situation of not being able to find the data. For light pollution data, many data open databases can be obtained directly, and we will collect and organize them for you later .
For example, if we want to collect economic data, we can collect the GDP of the year, or per capita income, etc., which can be collected broadly. Remember, you can't look for the dead truth, press one to collect it, it is easy to fall into the situation where you can't collect it. Therefore, it is enough to collect data and grasp the general direction. Here are some suggestions for you. For the economy, we can collect GDP; for land use, we can collect the area occupied by previous Olympic and Winter Olympics, and the total area of the city or the total area of the country for comparison. It should be difficult to collect people's satisfaction, but it can be replaced, that is, we can collect people's happiness index in the year of the Olympic Games; for tourism, we can collect fiscal tourism revenue, inbound population, etc.; for opportunities for improvement, we can consider government-related Policy-related data; international prestige can be replaced by some international rankings. In short, collect data without taking it for granted. Later, I will collect and share the data I mentioned, hoping to be helpful to everyone.
Data processing, it is impossible for us to use the collected data directly. The first thing to do is data preprocessing. We need to clean up the data first, that is, the data preprocessing stage. The processing of missing data varies from team to team. It is up to each team to decide whether to delete samples with missing data or modify them. For outliers, our first step must be the judgment of outliers, to judge what is an outlier, 3sigema principle, box diagram, etc., to define outliers.
The selection of indicators , for the selection of indicators, here we can first conduct research on problem 2, select indicators that are helpful to our problem research, and there are many methods for determining the weight of indicators, such as hierarchical analysis, principal components, entropy weight method, rank sum Ratio and so on are more or less the same and can vary from team to team.
Question 2. Consider Feasibility, Timelines for Implementation, and Impact of Potential Strategies on Your Metrics
For the three questions of question 2, I think that different models can be established to solve them. Here is just a suggestion. If you have different ideas, you can private message me to supplement.
For feasibility , I think we can build a predictive model, and for two different ideas, we build predictive models. (The selection of the prediction model can let a hundred flowers bloom. The simplest types of regression prediction in the prediction model, time series, neural network data, gray prediction, fuzzy prediction, etc. are all feasible. If you are not satisfied, you can also choose some advanced ones. Forecasting model, the corresponding code information will be shared later, etc.) To predict each data in question 1, you can also choose a kind of data as the dependent variable, and use the index in question 1 as the independent variable to predict it, Used to compare the quality of two ideas.
As for the timetable for implementation , I think it is possible to establish an optimization model, the objective function, without regard to the maximum value of the dependent variable, so that we can maximize the benefits at a certain time. The decision variable can be set as xi, which corresponds to different time schedules, or it can be the 0-1 variable of xij, which corresponds to different time schedules under the two ideas. Each team should have different opinions on the setting of decision variables. Here Just offering my suggestion for reference only. According to the different indicators of the constituency, set the corresponding constraints.
For the impact of potential policies on indicators , here we can consider it as a kind of sensitivity analysis, which can be analyzed for different potential policies. The potential policy here requires everyone to show their talents.
Question 3. Submit a one-page memorandum to the IOC describing your strategy and policy proposals
Question 3 is a typical problem of digital modeling. Write a non-technical article. What you need to pay attention to here is to write a memorandum to the Olympic Committee. You can search for the way the memo is written, and imitate some better, more exquisite ones, which look better. Fancy memos for parodying. In this way, for the 7-point review, it is easy to increase the judges' favor, and it is possible to even raise a level.
The last point is the attention of the data. We can find that for the YZ two questions in the spring competition, they are all questions with quite open results. We usually like this kind of questions with open results, the reason is that for this kind of questions, his answer must not be a fixed value, so as long as it is reasonable. If it is guaranteed to be reasonable, we need to read the literature roughly to understand the current situation. As long as the result is not outrageous, the judges cannot directly judge our paper wrong. Therefore, when we really can't find the data, or the data we find is not good, and the results of the code programming are not ideal, for this kind of open result topic, we should fabricate a data set, or fabricate a reasonable result It is understandable.
Finally, I wish everyone a smooth competition! ! ! ! ! ! !