Exponential Smoothing Forecasting Approach. Lecture # 11

This lecture is about the forecasting technique called Exponential Smoothing Approach.
We introduce a constant called smoothing constant alpha. And provide the fraction of the error to the future demand.
Ft = F(t-1) + a(A(t-1) - F(t-1))
Topics:
0:00 Introduction to exponential smoothing technique
2:29 Equation for exponential smoothing
5:50 Example solved using exponential smoothing

[@roshanmarode4679]

I like it. Thank you for such a wonderful explanation sir

[@bontu4018]

I have one question how to a form product dd for the least period is given below period dd period 1,2,3,4,5,6,7,8,9 dd 44,52,50,54,55,55,60,56,62 develop a linear trend equation ,predict the dd value for the next period and find the regression equation solution

[@karllocke4088]

What factors could you use to derive a suitable alpha, could you please do a lecture on alpha selection, many thanks

[@raishussein4046]

Just watching this 5 mins before exam thanks bro

[@technicalpeaks]

I hope your exam did go well.

[@midnightdishwasher1071]

hello is this example connected to the previous topic?

[@technicalpeaks]

Yes. This is another technique commonly used for forecasting. I made it separately because the concept is a little different from the rest!

[@ishratabdullah4764]

Sir would uh plz explain how uh got 144. Requesting uh

[@ishratabdullah4764]

Explanation understood 142+ .2×153_142 142+.2×11 142+2.2 144.2

[@pashtun-travels-uk]

Anyone know any thing about Simulation and modeling? Or MATLAB software?

[@technicalpeaks]

Do you want me to make MATLAB simulation for mathematical modeling video??

[@pashtun-travels-uk]

@@technicalpeaks create simple but appropriately organised spreadsheet models for complex problems, then use the spreadsheet to support traditional operational research techniques such as forecasting, optimisation and simulation describe the main categories of forecasting technique and their data requirements construct and interpret forecasts using smoothing methods and decompositional analysis calculate and interpret forecast errors to track accuracy of forecasts understand effective forecasting systems, basic principles of computer programming and VBA for developing macros within Excel

[@pashtun-travels-uk]

@@technicalpeaks please help me

[@pashtun-travels-uk]

Do you know these things which I wrote in message? Sir

[@technicalpeaks]

🌻🌻

[@a.j2376]

hi. I noticed online and in your lecture you use a different formula then my teacher did. I'm a little confused if there's a difference? We were told to use the following formula. I think its the Simple exponential smoothing formula: F(t+1)= a*Yt + (1-a)*ft where a= alpha F(t+1)= next period actual demand yt= observed values of series in period t then to find the error we subtract estimated and actual demand, and also find the error squared. Both ways provide the answer 144.2, but when do we use this method as opposed to your method shown? Do we use the method I mentioned when we need to find error and error squared only?

[@technicalpeaks]

Your professor formula and what I used are exactly the same formulas! They are just arranged differently and different names for same variables. F(t+1) = a×Yt + (1-a)×ft This is your formula. First change the names for variables according to mine ones. Variable = yours = mine Next period = F(t+1) = Ft Previous period Actual = Yt = F(t-1) Previous period forecast = ft = F(t-1) Now place my variables in your formula; F(t+1) = a×Yt + (1-a)×ft Ft = a×A(t-1) + (1-a)×F(t-1) Multiplying F(t-1) into the bracket Ft= a×A(t-1) + F(t-1) - a×F(t-1). Taking a common from first and third term. Ft= F(t-1) + a [ A(t-1) - F(t-1) ]. This is my formula. So, both are same!!! Hope I answered your question!! Take care and best wishes!!

[@a.j2376]

@@technicalpeaks thank you! I realized later that they both result in the same answer. Jazakhallah khair for the helpful videos! Got a perfect score on the test thanks to your video.

[@technicalpeaks]

I am Glad to be of help!!! Take care!! Best Wishes!!