The video describes using MS Excel to calculate seasonal indices for a very simple set of data that has seasonal fluctuations.
Пікірлер: 86
@sbhanuprakash93 жыл бұрын
Thanks for detail explanation. But may I know how we got the 4th year # as 570.(@5:00 min)
@aaronhei3119 жыл бұрын
Thank you for your wonderful and clear explanation!
@chatwithvarun4 жыл бұрын
Thanks so much. Very easily illustrated
@nidhichowdhary14732 жыл бұрын
Hi Piyush, can please calculate the fitted values using the model to calculate variance analysis and also how did you get total of 570 for year 4?
@suraj123ification8 жыл бұрын
Hello Mr. Piyush, Very nice video! I just have one question, though.. How will I calculate my seasonal indices when I do not have a complete set of data points? For example, if I am using monthly sales data for forecasting, and I have only 30 months of previous sales data, how will I calculate the seasonal indices? My data starts in Feb,2014 and ends in July,2017.
@Narduath10 жыл бұрын
This was extremely helpful, and served to clear up my confusions ahead of my Operations Management final. Thank you!
@piyushashah110 жыл бұрын
Am happy that the video helped you. Thanks for leaving this comment.
@lawrencenanagyan48910 жыл бұрын
where do you get the 570 from around the 5min mark?
@piyushashah110 жыл бұрын
Sorry for the late reply. I somehow missed your message. The 570 is a number I have assumed here. The yearly sale could have been obtained by any time series method. The yearly demand is not affected by seasons as all versions of seasons are inside the year and hence calculating that is easier.
@theknowledgekart4 жыл бұрын
he had made mistake..instead of 570 , 143.34 average of 1st quarter of all 4 years need to be taken and multiply it by seasonality factor of 1.0305 to get forecast of quarter 1 and so on for quarter 2, 3,4 .
@sanketnaikhe58192 жыл бұрын
Thank you sir
@abhishekarya0097 жыл бұрын
Nice video, Sir please make videos using same example for weighted moving average and exponential smoothing.
@josephnukonze39704 жыл бұрын
What would be the best method if the data does not show seasonality?
@ciai73178 ай бұрын
Where that 570 for year 4th come from?
@mayasilver65353 жыл бұрын
Do i have to predict year 4 or there is some calculation to get it ?
@theknowledgekart4 жыл бұрын
dear sir, thank u for the video..there is need of little correction in the tutorial.. instead of 570 , 143.34 average of 1st quarter of all 4 years need to be taken and multiply it by seasonality factor of 1.0305 to get forecast of quarter 1 and so on for quarter 2, 3,4 .
@piyushashah14 жыл бұрын
hello Rishu, thank you for this comment. The actual demand numbers already have the influence of seasons. So, averaging them and adding the effect of seasons again on them would not be valid. The ideal way is to use some method to find deseasonalized demand forecast and then add the effect of the season. I have used aggregate demand forecast from an unmentioned method as the deseasonalized demand. There can be many other ways to arrive at the deasonalized demand other than what I have done. Hope this helps.
@IAKhan-km4ph4 жыл бұрын
nice
@keranevelyne49872 жыл бұрын
How did you find the total for the fourth year (570)
@jeddturcotte51849 жыл бұрын
Thank you for your help! Was very informative.
@ricardoguzman7518 Жыл бұрын
This video was so good up until the 5 min mark. Year 4 being introduced with no explanation for the 570 demand......
@piyushashah1 Жыл бұрын
It is an assumed number that we could have obtained from regression or any other method.
@sasikumar3610Ай бұрын
how the expected requirement demand 570 and ave 142.5 calculated
@piyushashah1Ай бұрын
You would have to use some alternate method with yearly data and find the yearly demand (=570). And, then 540/4 (-142.5) is the deseanalized average quarterly demand.
@jumanjned58124 жыл бұрын
really thank you so much it helped me a lot for my MGT 300 final project, however i have a question that do we have to use a forecasting technique to get the predicted total demand for 2020?
@rachnasahu7838Ай бұрын
Hi, can we use the same method for calculating indices at weekly level throughout the year?
@piyushashah1Ай бұрын
Yes you may. But, it will get very complex like that. I would work at montly levels and then further disaggregate the monthly forecast into weeks. And, you could consider this app for monthly forecasting: zerohour.shinyapps.io/ETS_ARIMA_Prophet_monthly/
@afhdaily665 жыл бұрын
Thank you for very nice video. I am a beginner and autodidactive learner. It is very helpful
@wowwhydopplhavemyid11 ай бұрын
Watching your vids again! 10 years working in supply chain, but your videos always give a great reminder!
@piyushashah111 ай бұрын
Thanks for writing this comment Bryce.
@ankushzap8 жыл бұрын
what if I donot have 4th year total data 570. How to forecast then?
@piyushashah18 жыл бұрын
+ankushzap With this method you need that '570'. We can use some other method, maybe regression to get that.
@Sgenea8 жыл бұрын
Thank you for your video!
@jackthebeanstalkdelohio51463 жыл бұрын
For monthly seasonal pattern, how do an analyst calculate the average price change? Do they total all the gains, and get the average from that? Or do they total all the losses....... I am not sure. What is it if u don't mind me asking? Thx.
@tinatse199810 жыл бұрын
Many thanks for this video! Just dropped an email =)
@sfpodcastedge7430 Жыл бұрын
Is it possible to use the demand figures for year 3 multiplied by the seasonal indices to predict year 4 instead of assuming 570 as you did? I actually did use the excel triple smoothing model and got slightly higher forecast figure than these values..Thanks!
@agroasesor17 жыл бұрын
A big H E L L O from COLOMBIA. I do something like you.....related to milking production, but using percentage....I know you are laughing now but, is my way...and it works with cow´s milk production. I will use your method. G R A C I A S M I L !!!!
@asitkumarsadangi12955 жыл бұрын
Thank you so much. It helped me analyse data for my PG Thesis.
@smaranrai48983 жыл бұрын
Hi! can we find the seasonal indices for data that only have 1-year of data?
@piyushashah13 жыл бұрын
You can... But I won't trust the accuracy of such one year based indices.
@smaranrai48983 жыл бұрын
@@piyushashah1 Thank you for responding, but can you please tell me why wont you trust its accuracy? You see I am new to forecasting, and I wanted to know if a company only has past data for 1 year, how should it go about forecasting for future months mitigating the seasonal effects?
@piyushashah13 жыл бұрын
@@smaranrai4898 There could have been some random variation in a specific period. If you just have one year data, you will assume that the random variation is a part of seasonality. However, if you have multi-period data, such random variations tend to average out and we get more accurate seasonality indices.
@smaranrai48983 жыл бұрын
@@piyushashah1 Thank you so muchh piyush!! This makes things much clearer!!!
@anshulagarwal88314 жыл бұрын
Hi piyush, how you get the value of 570?
@piyushashah14 жыл бұрын
That is an assumed number Anshul. Typically we could use some form of time series of regression to arrive at it. I have a video on that.
@swapnildambhare2 жыл бұрын
570 year 4 it is immaginary?
@piyushashah12 жыл бұрын
As mentioned in the video, you can use some other method to forecast that.
@priyankakumarisinha59972 жыл бұрын
How this 570 came?
@piyushashah12 жыл бұрын
I have assumed that number for this video. We can get it from moving average, regression or any of the other methods explained in other videos.
@mushroomsavvy3 жыл бұрын
Thank you for your video. Can we first deseasonalize the demand data and do the forecast based on any of the forecasting methods such as exponential smoothing etc and then apply the seasonality index?
@marieelaineechano90858 жыл бұрын
thank you. this video is very helpful
@rosariodizon17538 жыл бұрын
hello Mister Pitush .. where did you get the 570 demand for year 4?
@piyushashah18 жыл бұрын
+Rosario Dizon That is assumed to be given. Maybe some other method was used to derive it.
@ricardoguzman7518 Жыл бұрын
@@piyushashah1 Why would someone try to figure out the forecast if that amount was already given?
@piyushashah1 Жыл бұрын
@@ricardoguzman7518 You would need monthly or quarterly demand forecasts for production or fullfilment. it is a normal process to disaggregate forecasts like this from some aggregate forecast.
@fernandesc236 жыл бұрын
Thank you! this was tremendously helpful :)
@piyushashah16 жыл бұрын
Do look at the other videos on forecasting as well.
@asimashahzadi98517 жыл бұрын
what is solution? when Anova result is not significant .
@piyushashah17 жыл бұрын
Asima Shahzadi sorry, can you elaborate on your question?
@sgteamx8 жыл бұрын
Hi jus wanna say thank you!
@rakschanna8 жыл бұрын
Excellent !!!! Thank you so much
@agnesmo85248 жыл бұрын
Mr. Piyush can I ask what book or theory references you used for this method?
@maryjohannaespera56438 жыл бұрын
i'm wondering how you can forecast daily volumes for christmas, using trends from past 3 years, as christmas day varies per year
@piyushashah18 жыл бұрын
I am not sure if you can use these methods to directly predict the demand for next Christmas. While you could use the Holt-Winter's model as a base, you would have to add contextual factors to account for current year.
@DianitaAB9 жыл бұрын
thanks!
@yanuaranabawahyuesa90499 жыл бұрын
Why not test the data from forecast eror ? How we know that forecast is well ?
@piyushashah19 жыл бұрын
Yanuar Anaba Wahyuesa We should check the Forecast for errrors. There is a separate video on the different ways to calculate the errors.
@swatikanchan86509 жыл бұрын
Hi Piyush, can you also show us how to deseasonalize the seasonality trends. Thanks
@piyushashah19 жыл бұрын
HI Swati, I'll try. Thanks a lot for the suggestion.
@akshitagulati97908 жыл бұрын
How would calculate seasonal indices using average percentage method?
@piyushashah18 жыл бұрын
Do you mean percentage of sales method? Why do you need to do that? Can you give me details please?
@Wachuka7 жыл бұрын
Very helpful. Thank you so much
@prgoddess10 жыл бұрын
Thank you! Very helpful!
@vaiataleone17524 жыл бұрын
how do we get the 570
@piyushashah14 жыл бұрын
That is an assumption for this video. In actual scenario we can get it from linear regression or some mechanics performed on yearly demand data.
@AndyRyanTX8 жыл бұрын
Thank you!!! You explain things so well and make it easy.
@piyushashah18 жыл бұрын
+Andy Ryan Thanks for leaving the comment. All good things are always simple!!
@razenndrajoshi41032 жыл бұрын
How to choose the correct forecast and do accuracy for retail? & How to do SKU level forecasting for the short term which includes seasonal items? I have sent mail. please check
@deepsammanna7 жыл бұрын
Sir how do you model the croston method for intermittent demands?
@piyushashah17 жыл бұрын
There are pdfs online that discuss Croston method. For example this: courses.edx.org/asset-v1:MITx+CTL.SC1x_1+2T2015+type@asset+block/w4l2_NewIntermittentProducts_ANNOTATED_FINAL.pdf. I have never used this method am not sure how exactly to use this.
@gorecki46129 жыл бұрын
Great click noise
@piyushashah19 жыл бұрын
Daniel Burke Yes, and sorry. But the free tool I have has this limitation. I hope I can get a better method or a tool.
@svenskfox9 жыл бұрын
Really helpful, thanks a lot!
@piyushashah19 жыл бұрын
Thanks for leaving this comment. Happy that the video was useful to you. Do also check the video on Measuring Forecasting Accuracy.
@last123login9 жыл бұрын
Thanks for sharing
@piyushashah19 жыл бұрын
Thanks for leaving the comment Aay...there are other forecasting videos as well, do have a look.