First question
It is required to analyze the distribution patterns and interrelationships of sales volume of various vegetable categories and single products. This problem can be analyzed from three angles.
- 1) Sales volume distribution of various types of vegetables, and the relationship between vegetable types and sales volume;
- 2) Monthly distribution of sales volume of various types of vegetables, and correlation between sales volume of various types of vegetables and months;
- 3) The distribution of sales time of various types of vegetables, the correlation between sales time and the amount of returns.
- For each angle, box plots and line plots are first used to express the distribution of special diagnoses, and then the Kolmogorov-Smirnov distribution test is performed, and finally the corresponding correlation analysis method is selected for calculation.
Second question
- First, we need to analyze the relationship between the total sales volume of various types of vegetables and cost-plus pricing. In essence, we need to find the relationship between the total sales volume and cost profit margin. For this problem, an integrated fitting model is established to fit a polynomial function for each vegetable item to characterize the relationship between sales volume and cost profit margin.
- Secondly, it is necessary to give the total daily replenishment volume and pricing strategy of each vegetable category in the next week. After obtaining the relationship model between sales volume and cost profit margin, the sales volume can be calculated through the cost profit margin. Therefore, a mixed integer linear programming model (MILP) was constructed with the goal of maximizing supermarket profits and using the cost profit rate of each vegetable item as the decision variable.
- For the maximum sales constraint among the constraints, vegetable demand is predicted based on the ARIMA algorithm, and the upper 95% value of the confidence interval is taken as the maximum sales constraint. And solve the MILP model based on genetic algorithm (GA). For the discount strategy, an integrated model is first used to calculate the relationship between discount ratio and sales volume, and then a dynamic adjustment model is constructed to adjust prices based on actual sales.
Question 3
In essence, it adds restrictions on sales space and minimum display quantity on the basis of question 2.
To this end, the corresponding restrictions are adjusted on the basis of the MILP in question 2 to limit the sales space and minimum display quantity. In the solution phase, a 0-encoding mechanism is designed to adapt to the sales space constraints, and an unsolvable repair mechanism is designed to maintain the legitimacy of the chromosome. Use the dynamic adjustment model constructed in question 2 to make price adjustment decisions.
Question 4
The collection of relevant data needs to be designed. This question can be analyzed from the perspectives of "whether there is a member discount", "the best storage time of each product", and "reasons for returns".
Students who need ideas and subsequent complete codes can follow me~