The author discusses the strategies used to build predictive models, focusing on time series forecasting, regression analysis, and time series forecasting with covariates. Time series forecasting analyzes data collected over time to predict future values or trends based on historical patterns. Regression analysis establishes a mathematical relationship between input variables and an output variable, quantifying their relationships. Time series forecasting with covariates incorporates additional independent variables to improve accuracy, especially when external factors have a significant influence. The author acknowledges that regression analysis and time series forecasting with covariates require careful feature engineering and continuous updates. Other techniques like Facebook Prophet and deep learning models can also be used for predictive modeling. The author highlights the importance of understanding these strategies to unlock the power of predictive models.
https://www.amorphousdata.com/blog/time-series-vs-regression