Applied Time Series Analysis and Forecasting 时间序列分析和预测 (MAV557)

MAV557 Applied Time Series Analysis and Forecasting focuses on using data to reduce uncertainty in decision-making through effective forecasting techniques. Students will explore key properties of time series data, such as seasonality and trends, and learn to construct predictive models using time series regressions and tools like the ARIMA model. Designed for industry practitioners, the course emphasizes practical applications and real-world use cases of forecasting in business and policy-making. By utilizing both scripting-based (e.g., Python) and non-scripting-based software, students will gain hands-on experience in deploying forecasting results, enabling them to apply these techniques to scenario planning, budgeting, and data-driven strategies in various industries. MAV557时间序列分析与预测专注于通过有效的预测技术利用数据来减少决策中的不确定性。学生将探索时间序列数据的关键特性，如季节性和趋势，并学习使用时间序列回归和ARIMA模型等工具构建预测模型。该课程面向行业从业者，强调预测在商业和政策制定中的实际应用和真实案例。通过使用脚本编程（例如Python）和非脚本编程软件，学生将获得部署预测结果的实践经验，使他们能够将这些技术应用于各种行业中的情景规划、预算和数据驱动的战略决策。

• Basic properties of time series data 时间序列数据的基本特性
• Time series visualisations 时间序列可视化
• Overview of linear regression 线性回归概述
• Practical issues in time series regression 时间序列回归中的实际问题
• Autoregressive models 自回归模型
• Predictor selection criteria 预测变量选择标准
• Modelling trends 趋势建模
• Modelling seasonality 季节性建模
• Stationarity and differencing 平稳性和差分
• Moving average and autoregressive models 移动平均和自回归模型
• ARIMA models ARIMA模型
• Estimation and model selection of ARIMA models ARIMA模型的估计和模型选择

• Design a forecasting model appropriate for the available historical data设计适合可用历史数据的预测模型
• Assess the properties of time series data评估时间序列数据的特性
• Appraise the suitability of various forecasting frameworks评价各种预测框架的适用性
• Select the suitable forecasting framework for the business problem为业务问题选择合适的预测框架
• Prepare data for forecasting为预测准备数据
• Construct a forecasting model 构建预测模型