forecasting

Basic Concept: Forecasting involves using past data to predict future outcomes. This can include analyzing trends over time and identifying relationships between variables.
Key Example: Analyzing club food purchases over various time frames (e.g., comparing this year with the last year) to foresee future trends.

Definition: A method that uses historical data trends to predict future values, focusing on patterns over time.
Purpose: Identifies consistent trends (e.g., increase or decrease) to make predictions about future values.
Application: Example of forecasting dollar store purchases where an increase might indicate socioeconomic changes.

Definition: Involves establishing relationships between different variables.
Example: Understanding that dollar store purchases are usually higher among individuals with lower incomes.
Regression Analysis Class: Forecasting involves creating a regression model that relates income levels to purchasing behavior.
Leading Indicators: Certain purchases can predict economic events, e.g., increased dollar store purchases might foreshadow a recession.
Animal Behavior: Predicting earthquakes based on unusual animal behavior is a metaphor for how some variables can indicate extreme events.

Definition: A concept where collective opinions may provide accurate forecasts.
Prediction Markets: Utilize crowdsourcing to gauge future outcomes based on the betting behavior of participants.

Definition: Statistical models used to capture relationships among economic variables to help forecast future outcomes.
Example: Analyzing quarterly data on airline seat prices and sales volume.
Regression Technique:
- Estimate the relationship between price and quantity to forecast future sales changes based on price adjustments.
- Equation: If price decreases by 10%, the corresponding change in quantity can be calculated based on regression results.

Assessment of Forecasts: Evaluating the accuracy of forecasting models through various error metrics.
Types of Forecast Errors:
- Mean Absolute Error (MAE): Average of absolute errors.
- Root Mean Square Error (RMSE): Measures the square root of the average squared differences between forecasted and actual values, which penalizes larger errors more heavily.
Importance: Helps in refining models to improve predictive accuracy.

Dataset Example: Quarterly data on toy sales, where sales patterns can be analyzed seasonally (winter, spring, summer, fall).
Regression: Use simple linear regression to identify trends, seasonal adjustments using dummy variables for different seasons can refine predictions.
Model Differences: Comparing various models like quadratic vs. linear to discover which provides more explanatory power.

Construction of Dummy Variables: Creating binary (0/1) variables to indicate seasons, e.g., set 'winter' to 1 if applicable, and 0 otherwise.
Significant Results: Analyze t-statistics from regression to determine statistically significant predictors.

Trailing Data Method: Using previous sales as predictors for future sales.
Period Lagging: Analysis shows importance of data from multiple preceding periods, e.g., quarterly sales lag impacts next quarter's predictions.

Interaction Terms: Combining time trends with seasonal indicators may reveal how seasonal impacts evolve over time.
Data Construction: Multiply time indicators with seasonality dummy variables to test interaction significance.

Complexity in Forecasting: It’s crucial to assess a wide array of factors when making predications, including economic changes and seasonal variations.
Model Flexibility: Different models can yield varying degrees of accuracy, requiring continuous validation and adjustment based on external conditions.
Continuous Improvement: Regularly assess forecast models to ensure that they remain accurate and relevant to changing economic landscapes.