Sign up now
to enroll in courses, follow best educators, interact with the community and track your progress.
Forecasting Error (in Hindi)
490 plays

This lesson throws light on forecast error

Harshit Aggarwal
Cleared UPSC ESE twice with Rank 63 and 90 in mechanical engg. Got 99 percentile in GATE. Cracked ONGC, BHEL,ISRO, SAIL, GAIL successfully

Unacademy user
sir at 6:30 n=3 kaise pata chala

  2. ABOUT ME Graduated from NIT Nagpur in 2008 Cleared Engineering Services Examination (ESE-UPSC) Exam Got the offer letter from most of the Maharatna and Navratna Companies Cleared GATE Exam Rate, Review, Recommend, Share Follow me on Unacademy at: aggarwal

  3. MEAN ABSOLUTE DEVIATION (MAD) Forecast error is used to find the pattern which may regulate our future production. The error should be minimum as far as possible and the most generally used techniques to find forecast error are: 1. 90 120 30 2. 130 100 +30 MAD 1-301+30 -60 It tells the absolute magnitude of forecast error for certain number of periods (without considering sign)

  4. MEAN FORECAST ERROR (MEE) OR BIAS MFE or Biasyn DI-Fi It only tells the direction of forecast error and shows any tendency of over forecast or under forecast. Positive bias means under estimated forecasting and negative bias means over estimated forecasting. Running sum forecast error (RSFE) i-1Di F and RSFE BiasL

  5. MEAN SQUARE ERROR (MSE) i-1 It is used to plot control chart for forecast error and nowadays it is the most used one.

  6. MEAN ABSOLUTE PERCENTAGE ERROR (MAPE) Di-Fi It is the mean of percentage error or deviation as compared to actual demand and it is used to compute error with respect to demand because there is difference between 50 out of 100 and 50 out of 1000

  7. TRACKING SIGNAL (TS) RSFE TS= MAD It tells how well the forecast is predicting the actual value. A value of 'zero' is ideal but 4 or 5 is the acceptable range.

  8. QUESTION What are moving average and exponential smoothing models for forecasting? A dealership for Honda city cars sells a particular model of the car in various months of the year. Using the moving average method, find the exponential smoothing forecast for the month of October 2010. Take exponential smoothing constant as 0.2: Jan Feb. March 80 65 90 70 80 100 cars 85 65 75 2010 2010 2010 2010 cars cars cars cars cars rl May June July Aug Sept. 2010 2010 2010 2010 2010 cars cars cars

  9. SOLUTION Months Jan Feb March Sells cars 80 65 90 70 80 100 85 60 75 Forecast demand (n-3 Forecast of oct. by exponential smoothing method + oc (sep.sep. 80+65+90)/3-78.33 65+90+70)/3-75 90+70+80)/3-80 70+80+100)/3-83.33 80+100+85)/3-88.33 100+85+60)/3-81.67 o 0.2 F 73.33 Dspt 75 Foct 81.67 +0.2 (75-81.67) oct 80.33 T1 sep Ma June July Au Se 81 Forecast for the month of October using moving average Sep oct 8060+75 71.67