MODULE 8 PART 2

• Quantitative


  Relating to or involving the measurement or analysis of quantities or amounts
• Forecasting


  The process of making predictions of the future based on past and present data and most commonly by analysis of trends
• Overview of quantitative methods

  • Time series models
  • Associative models
• Time series models


  • Forecast based only on past values
  • Assumes that factors influencing past and present will continue influence in future
  • Future is a function of the past
• Time series


  A set of evenly spaced numerical data obtained by observing response variable at regular time periods (weekly, monthly, quarterly and so on)

  
• Additive model


  Incremental in effect, Yi = Ti + Si + Ci + Ri
• Multiplicative model


  Exponential in effect, Yi = Ti · Si · Ci · Ri
• Time series components

  • Trend
  • Seasonal
  • Cyclical
  • Random
• Naive approach


  Simplest approach; Assumes demand in next period is the same as demand in most recent period OR follows the same pattern based on a defined component
• Moving averages


  Uses an average of the n most recent periods of historical data to forecast the next period

  
• Exponential smoothing


  Uses a smoothing constant (α) - the higher the value, the more sensitive your forecast becomes to more recent periods
• Trend projection


  Technique developed to correct the failure of the simple exponential smoothing to respond to trends by adding trend adjustments
• Time-series models are analysis based on the idea that the history of occurrences over time can be used to predict the future
• Associative models


  • Tries to understand the system underlying and surrounding the item being forecasted (e.g., sales maybe affected by advertising, quality, & competitors)
• Associative models

  • Regression analysis
  • Econometric model
  • Input/output models
  • Leading indicators
  • Box Jenkins technique
  • Shiskin time series
• Simple moving average


  Uses an average of the n most recent periods of historical data to forecast the next period

  
• Weighted moving average


  Used when trend is present; Assumption is that older data usually less important, which means they are assigned smaller weights
• Disadvantages of moving average methods

  • Requires much historical data
  • Increasing n smooths out fluctuations, makes forecast less sensitive to real changes
  • Does not forecast trend well
  • Weights dictate the sensitivity to changes in trends, cycles, and seasons
• Evaluating the Smoothing Constant (α) affects the reaction rate of the forecasted values to the changes experienced by the historical demand data
• Exponential smoothing with trend adjustment


  Technique developed to correct the failure of the simple exponential smoothing to respond to trends by adding trend adjustments
• Trend Component


  represented as the long term movement in a time series, absent of any effect of calendar events or irregular factors
• Seasonal Component


  represented by regular up or down fluctuations caused by systematic calendar related influences such as weather or natural conditions, culture, customs, or yearly events
• Cyclical Component


  Events that also happen in a recurring manner. Contrary to seasons, cycles are not calendar related or have a specific timing and can last more than one calendar year.
• Random Component


  shifts in the demand that are unpredictable, erratic, unsystematic and irregular