Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.



Published on

Published in: Business, Technology
  • Be the first to comment

  • Be the first to like this


  1. 1. Forecasting
  2. 2. Why Forecast ? To Plan For Demand Its not a goal
  3. 3. Vocabulary of Time Series Trend – gradual upward or downward movement of the data over time Seasonality – demand fluctuation pattern above and below the trend line repeating at certain points of time ( not limited to years) example – business at the bank with seasonal effects around mid and end of month paydays Cycles – patterns that occur in the data every several years (reflect business cycles) Random Variations ( Special Causes )– chance variation in the data – no pattern – like error in statistical model
  4. 4. Graph of a Time Series
  5. 5. Analyse data Univariate A No Yes Plot Data into I-MR Chart Huge Variations No Correct For Special Causes Decompose Stationary ? Trend & Seasonality Single Exponential Smoothening Double Exponential Smoothening Holt -Winters Yes No No Stable after Checking for n/2 vaues/ errors? Forecast Value Regression MAPE ,MAD, MSE Should be very good pointers Yes Yes Yes Yes No
  6. 6. Illustration Of Errors
  7. 7. Measures of Forecast Error <ul><li>Bias - The arithmetic sum of the errors </li></ul><ul><li>MAD - Mean Absolute Deviation </li></ul><ul><li>MAPE – Mean Absolute Percentage Error </li></ul><ul><li>Mean Square Error (MSE) - Similar to simple sample variance </li></ul><ul><li>Standard Error - Standard deviation of the sampling distribution (the square </li></ul><ul><li>root of the MSE) </li></ul><ul><li>Bias, MAD, and MAPE - typically </li></ul><ul><li>used for time series </li></ul>Minitab Gives all 
  8. 9. THANK YOU !!!! DIrect LIase CONfirm