Mathematical Modeling Competition Projects

This repository contains award-winning mathematical modeling competition projects, showcasing comprehensive capabilities in data analysis, predictive modeling, and optimization decision-making.

📁 Project Overview

This repository includes two award-winning mathematical modeling competition projects:

1. TSLA - Vegetable Price and Sales Forecasting with Pricing Optimization

Award: 🏆 National Second Prize

Project Description:

• Vegetable sales forecasting and pricing strategy optimization based on time series analysis and genetic algorithms
• Study of sales distribution, seasonal variations, and complementary-substitutive relationships among vegetable categories
• Proposing optimal ordering strategies to maximize profit and balance category supply

Core Technologies:

• Time Series Decomposition (Seasonal Decompose)
• Multiple Linear Regression
• Genetic Algorithm Optimization
• Cluster Analysis
• Correlation Measurement

Key Achievements:

• Vegetable category sales trend prediction model
• Restocking quantity and pricing strategy optimization algorithm
• Sensitivity analysis and robustness validation

---

2. PUMC - Dynamic Tennis Match Prediction Model

Award: 🏆 MCM/ICM Honorable Mention

Project Description:

• Real-time tennis match prediction system based on machine learning and dynamic modeling
• Construction of dynamic momentum indicators to evaluate match trends
• Multi-dimensional feature fusion for match outcome prediction

Core Technologies:

• Support Vector Machine (SVM)
• XGBoost Classifier
• Dynamic Feature Engineering
• Grey Relational Analysis
• SHAP Interpretability Analysis

Key Achievements:

• Dynamic match momentum evaluation model
• Real-time score and situation prediction algorithm
• Feature importance visualization analysis

---

🚀 Quick Start

Requirements

• Python 3.8+
• Jupyter Notebook
• Main dependencies:

  pandas
  numpy
  matplotlib
  seaborn
  scikit-learn
  statsmodels
  xgboost
  shap

Installation Steps

1. Clone the repository

git clone https://github.com/LZHMS/MathematicalModeling.git
cd MathematicalModeling

2. Install dependencies

pip install -r requirements.txt

3. Launch Jupyter Notebook

jupyter notebook

---

📊 Project Structure

MathematicalModeling/
├── TSLA/                      # Vegetable Sales Prediction Project
│   ├── data/                  # Data files
│   ├── images/                # Visualization results
│   ├── P1.ipynb              # Data processing and analysis
│   ├── P2-P3.ipynb           # Modeling and optimization
│   └── README.md             # Project documentation
│
├── PUMC/                      # Tennis Match Prediction Project
│   ├── data/                  # Data files
│   ├── images/                # Visualization results
│   ├── codes.ipynb           # Complete code
│   └── README.md             # Project documentation
│
└── README.md                  # This file

---

🎯 Technical Highlights

Data Processing

• Large-scale data cleaning and preprocessing
• Missing value imputation and outlier detection
• Feature standardization and normalization

Modeling Methods

• Time Series Analysis: Seasonal decomposition, trend forecasting
• Machine Learning: SVM, XGBoost, PCA dimensionality reduction
• Optimization Algorithms: Genetic algorithm, constrained optimization
• Statistical Analysis: Correlation analysis, cluster analysis

Visualization

• Dynamic trend charts
• Heatmaps and scatter matrix
• Feature importance visualization
• Hierarchical clustering dendrograms

---

📈 Key Results

TSLA Project

• ✅ Accurate prediction of vegetable category sales trends
• ✅ Identification of complementary and substitutive relationships between categories
• ✅ Optimal pricing and restocking strategies
• ✅ Model robustness validation

PUMC Project

• ✅ Real-time dynamic match momentum assessment
• ✅ High-precision score prediction (accuracy > 85%)
• ✅ Interpretable feature importance analysis
• ✅ Multi-level match outcome prediction

---

👥 Contributors

TSLA Project Team

• Zhihao Li
• Pai Lin
• Kaida Huang

PUMC Project Team

• Zhihao Li
• Changrong You
• Zhihan Liu

---

⭐ If this project helps you, please give us a Star!