Promedio Móvil Doble

Double Moving Average Forecasting Technique

Introduction to Double Moving Average

• The double moving average technique is introduced as a method for forecasting future periods based on historical data, specifically enhancing the capabilities of simple moving averages.

• Unlike simple moving averages that can only predict the immediate next period, double moving averages allow for forecasts extending beyond one period, such as predicting data for April through August.

Steps in Calculating Double Moving Average

• The first step involves calculating the simple moving average for 'n' orders, where 'n' represents the number of averages derived from sales or demand data.

• Next, select the forecast with the lowest average error. This process was elaborated in a previous video discussing how to calculate this error.

• Finally, the methodology requires recalculating using two types of adjustments: trend adjustment and empirical adjustment. Formulas are provided to facilitate these calculations.

Understanding Key Formulas

• A polynomial geographic formula is referenced (y = a + bx), where 'x' denotes additional periods to estimate. This formula is foundational in linear regression analysis.

• In this context, 'a' represents the slope while 'b' signifies the intercept; both are crucial for determining expected sales or demand over projected periods.

Practical Application Using Excel

• The objective is to find an appropriate slope based on historical data to project future values accurately against time versus expected demand or sales.

• Previous modeling efforts identified three types of genes (n3, n4, n5), revealing that an average error of 35 points indicates that n = 3 yields optimal results for forecasting accuracy.

Data Preparation and Analysis

• Historical data from 2018 will be utilized since it provides a reliable basis for making forecasts without incorporating erroneous predictions.

• A new worksheet will be created in Excel labeled "Simple Moving Average" to differentiate between various methods used in forecasting.

Implementing Double Moving Average Formulae

• The double moving average calculation involves applying formulas iteratively; it's essential to note that this method requires averaging twice as indicated by its name.

• To achieve accurate projections, specific formulas must be brought into Excel from prior presentations which detail how adjustments should be made based on empirical and trend analyses.

Forecasting Techniques Using Moving Averages

Introduction to Moving Averages

• The discussion begins with the introduction of two formulas necessary for forecasting, emphasizing the need for a simple moving average (SMA) and a double moving average (DMA).

• Historical data from 2009 to 2017 is used as a basis for predictions, with the final forecast being utilized in calculating the SMA.

Calculation of Simple and Double Moving Averages

• The SMA is calculated using historical data, specifically noting that it equals 37 based on the last predicted value.

• For DMA, similar calculations are performed; however, limitations arise due to insufficient prior data points needed for accurate forecasting.

Limitations in Forecasting

• The speaker highlights challenges in predicting future values when there are not enough historical data points available. For instance, only three previous years' data can be used for certain forecasts.

• The DMA result is noted as 31.44, which will be essential for further calculations.

Formula Application and Adjustments

• The formula application involves multiplying two times the SMA minus the DMA. This step requires careful attention to ensure accuracy in calculations.

• Emphasis is placed on understanding how constants and variables interact within the formula structure during calculations.

Practical Example: Delta Company Forecasting

• An example involving Delta Company illustrates how to forecast demand for 2018 using historical data up until 2017.

• Challenges are discussed regarding forecasting beyond immediate next periods (e.g., predicting for 2021 or 2024), highlighting limitations of current techniques.

Determining Future Period Values

• To predict future years like 2021 and 2024, it's crucial to establish 'x', representing subsequent periods after historical data.

Forecasting Demand for 2018 and Beyond

Introduction to Data Analysis

• The speaker introduces data that will be used for reference in forecasting demand, emphasizing the importance of having accurate historical data.

Calculating Forecasts

• The process begins with determining equivalences for the year 2018, using a formula involving multiplication of variables (b and x).

• The forecast for 2018 is calculated as 42.56 plus a product of b and x, indicating a structured approach to demand prediction.

Trend Analysis

• A trend analysis shows an increase in production from the year 2000 to 2018, predicting a demand of 48 units based on historical growth patterns.

Extending Predictions

• The speaker discusses extending forecasts into future years (e.g., 2019), suggesting that trends indicate an increase to 53 units.

• Emphasis is placed on maintaining consistency in moving formulas while adjusting predictions for subsequent years.

Final Forecast Results

• For the year 2021, the double moving average technique predicts sales of 64 units; by contrast, projections for 2024 suggest an increase to 81 units.

Visualizing Data Trends

• The speaker plans to create a graph using collected data to visualize demand trends effectively.

• A suitable graph is selected to illustrate data behavior over time, highlighting positive trends through graphical representation.

Adding Trend Lines

• A trend line is added to the graph, showing how projected values align with historical data points and confirming upward trends in demand.

Interpretation of Results

• The analysis indicates that if projections fall below expected levels (intercepting at y-axis), it may signal declining product relevance or market exit.

Conclusion on Demand Trends