Scribd Logo
Upload

Welcome to Scribd!

  • 67 views

    Linear Trend Projection: Deris & Dimayuga BSA 2

    Uploaded by

    natalie clyde mates
    The document describes using linear regression to forecast a time series with a linear trend, specifically bicycle sales data from the past 10 years. It shows calculating the linear trend line equation Yt = b0 + b1t that best fits the data by minimizing the sum of squared errors. This linear trend line is then used to forecast future bicycle sales, predicting a value of 32,500 bikes for next year. The accuracy of this linear regression forecasting method is evaluated by calculating the forecast errors and summed squared errors for the bicycle sales time series data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Linear Trend Projection: Deris & Dimayuga BSA 2

    Uploaded by

    natalie clyde mates
    67 views16 pages
    The document describes using linear regression to forecast a time series with a linear trend, specifically bicycle sales data from the past 10 years. It shows calculating the linear trend line equation Yt = b0 + b1t that best fits the data by minimizing the sum of squared errors. This linear trend line is then used to forecast future bicycle sales, predicting a value of 32,500 bikes for next year. The accuracy of this linear regression forecasting method is evaluated by calculating the forecast errors and summed squared errors for the bicycle sales time series data.

    Original Title

    15.4-LINEAR-TREND-PROJECTION

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document describes using linear regression to forecast a time series with a linear trend, specifically bicycle sales data from the past 10 years. It shows calculating the linear trend line equation Yt = b0 + b1t that best fits the data by minimizing the sum of squared errors. This linear trend line is then used to forecast future bicycle sales, predicting a value of 32,500 bikes for next year. The accuracy of this linear regression forecasting method is evaluated by calculating the forecast errors and summed squared errors for the bicycle sales time series data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    67 views16 pages

    Linear Trend Projection: Deris & Dimayuga BSA 2

    Uploaded by

    natalie clyde mates
    The document describes using linear regression to forecast a time series with a linear trend, specifically bicycle sales data from the past 10 years. It shows calculating the linear trend line equation Yt = b0 + b1t that best fits the data by minimizing the sum of squared errors. This linear trend line is then used to forecast future bicycle sales, predicting a value of 32,500 bikes for next year. The accuracy of this linear regression forecasting method is evaluated by calculating the forecast errors and summed squared errors for the bicycle sales time series data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 16

    Linear Trend

    Projection

    Deris & Dimayuga BSA 2


    In this section we present forecasting methods that are
    appropriate for time series exhibiting trend patterns. Here
    we show how regression analysis may be used to
    forecasting a time series with a linear trend. In section 15.1
    we used bicycle sales time series in table 15.3 and figure
    15.3 to illustrate a time series with a trend pattern. Let us
    now use this time series to illustrate how regression
    analysis can be used to forecast a time series with a linear
    trend. The data for the bicycle time series are repeated in
    15.11 and figure 15.9

    Although the time series plot in figure 15.9 shows some up


    and down movement over the past 10 years, we might
    agree that the linear trend line show in figure 15.10
    provides a reasonable approximation of the long-run
    movement in the series. We can use regression analysis to
    develop such a linear trend line for the bicycle sales time
    series.
    Table 15.11 BICYCLE SALES TIME SERIES

    Year Sales (1000s)


    1 21.6
    2 22.9
    3 25.5
    4 21.9
    5 23.9
    6 27.5
    7 31.5
    8 29.7
    9 28.6
    10 31.4
    Figure 15.9 BICYCLE SALES TIME SERIES
    PLOT
    Figure 15.10 TREND REPRESENTED BY A
    LINEAR FUNCTION FOR THE BICYCLE
    SALES TIME SERIES
    In regression analysis we use known values of variables to
    estimate the relationship between one variable (called the
    dependent variable) and one or more other related variables
    (called independent variables.) This relationship is usually
    found in a manner that minimizes the sum of squared errors
    (and so also minimizes the MSE). With this relationship we
    can then use values of the independent variables to estimate
    the associated value of the dependent variable. When we
    estimate a linear relationship between the dependent
    variable (which is usually denoted as y) and a single
    independent variable (which is usually denoted as x), this
    referred to as simple linear regression. Estimating the
    relationship between the dependent variable and a single
    independent variable requires that we find the values of
    parameters b0 and b1 for the straight line y = bo + b1x.
    Because our use of simple linear regression analysis
    yields the linear relationship between the independent
    variable and the dependent variable that minimizes the
    MSE, we can use this approach to find a best-fitting line
    to a set of data that exhibits a linear trend. In finding a
    linear trend, the variable to be forecasted (Y1, the actual
    value of the time series in period t) is the dependent
    variable and the trend variable (time period t) is the
    independent variable. We will use the following notation
    for our linear trend line.
    t = the time period
    Yt = linear trend
    forecast in period t (i.e., the estimated value of Yt in period t ) bo = the Y-
    intercept of the linear trendline
    b1 = the slope of the linear trendline

    LINEAR TREND LINE


    In equation (15.10) the time variable begins at t = 1
    corresponding to the first time series observation (Year 1
    for the bicycle sales time series) and continues until t = n
    corresponding to the most recent time series observation
    (Year 10 for the bicycle sales time series). Thus, for the
    bicycle sales time series t = 1 corresponds to the oldest
    time series value and t = 10 corresponds to the most recent
    year. Calculus may be used to show that the equations
    given below for b0 and b1 yield the line that minimizes the
    MSE. The equations for computing the values of b0 and b1
    are
    t = the time period
    Yt = actual value of
    the time series in period t
    n = number of periods in the time series
    Y = average
    value of the time series; that is, ---------->> t =
    mean value of t; that is, --->>
    Let us calculate b0 and b1 for the bicycle data in Table 15.11; the
    intermediate summary calculations necessary for computing the values
    of b0 and b1 are
    t Yt
    tYt
    1 21.6
    21.6
    2 22.9
    45.8 3
    25.5
    76.5 4
    21.9
    87.6 5
    23.9
    119.5 6
    27.5 165.0
    7 31.5
    And the final calculation of the values of b0 and b1 are
    is the regression equation for the linear trend component for the
    bicycle sales time series. The slope of 1.1 in this trend equation
    indicates that over the past 10 years, the firm has experienced an
    average growth in sales of about 1100 units per year. If we assume
    that the past 10- year trend in sales is a good indicator for the future,
    we can use equation (15.13) to project the trend component of the
    time series. For example, substituting t = 11 into equation (15.13)
    yields next year’s trend projection,
    Thus, the linear trend model yields a sales forecast of 32,500 bicycles
    for the next year. Table 15.12 shows the computation of the
    minimized sum of squared errors for the bicycle sales time series. As
    previously noted, minimizing sum of squared errors also minimizes
    the commonly used measure of accuracy, MSE. For the bicycle sales
    time series,
    WEEKS SALES FORECAST FORECAST SQUARED
    (1000S) ^ ERROR FORECAST
    Yt ERROR
    Yt
    TABLE 15.12 1 21.6 21.5 0.1 0.01

    2 22.9 22.6 0.3 0.09

    SUMMARY OF 3 25.5 23.7 1.8 3.24


    THE LINEAR 4 21.9 24.8 -2.9 8.41
    TREND 5 23.9 25.9 -2.0 4.00
    FORECASTS AND 6 27.5 27.0 0.5 0.25
    FORECAST 7 31.5 28.1 3.4 11.56
    ERRORS FOR THE
    8 29.7 29.2 0.5 0.25
    BICYCLE SALES
    9 28.6 30.3 -1.7 2.89
    TIME SERIES
    10 31.4 31.4 0 0
    _________

    Total 30.70
    THANK YOU

    You might also like