Hi, Thank you so much for the comprehensive training on regression techniques. I have a question. If a data set has both seasonality (quarterly sales) and trend, but the trend is expected to change after a certain period, how do I accommodate that using your techniques? The second question is if the data has more than one seasonality (more than one-quarter of low/high sales), will the technique work efficiently? Please let me know. Thanks.
Thank you Mike. Stellar job on your entire body of work. Regarding Categorical Variables: At 9:53 - 10:22, you state: "We're going to have to add 1, 2, 3 extra columns to our dataset. Why 3? Because if you have whatever number of categorical variables, you have one less column than there are variables." This seems to beget the question, why, "if you have whatever number of categorical variables" do you only need "one less column than there are variables"? What is the reason that one---in this case Q4---is not included in the same manner as the others, and why is it Q4 rather than one of the other 3? Perhaps this is related to your statement at 11:46 that: "The slope for quarter 4 will already be incorporated into the Y-intercept." How so? Why Q4? Could this use further explanation? Thanks. Best regards.
I still don’t get this. Have you understood it by now and could you please explain? Could the creator of the video maybe please elaborate on this topic since it seems like multiple people are struggling to grasp this concept. Thank you!
You choose X to be time, however other variable could be choose too. For an example price elasticity of demand price could be use as X variable, right ?
Hi Sir, your videos r very helpful. Just a quick question, how you determine when you will use moving average and liner regression? Based on MSR? Thanks
Good video. But just want to confirm: in example 3, the p-value for Qtr3 is 0.07 which indicates insignificance of Qtr3 coefficients. That needs to be fixed right?
Is there a particular reason for using the Catigorical column formulas you did at about 12:35?... =VLOOKUP(C4,$O$4:$R$7,MATCH(I$3,$O$3:$R$3,0),0) It's an instructive formula, but wouldn't it have served equally well to simply name the Categorical Columns "Q1, Q2, & Q3" and then used =IF(H$1=$C4,1,0) to compare each Catigorical column header back to the Quarter in Column C? Maybe I'm missing some additional subtle consideration here. Thanks. Best Regards.
Hi. What is the difference for calculating the slopes for each Quarter rather than getting the linear trend and averaging a percentage between the quarters from the historical data, so the four Quarter averages would be 100%. Wouldn't it create a very similar forecast?
Hi which of your template can help me reviewing within mfg capacity[PO +/- [% FCST]] . i am trying to see if i can create order report that tied up mfg capacity , FCST, Serves level , SS ...
Hi, Can you tell me how did you manage to show the formula in I column and values in G column simultaneously at +3:07. I'm looking for the same thing for very long time but no success yet. Thanks
Satish, You could figure this out simply by downloading the provided Excel file from the Excel Is Fun website and examining the cell contents. Nonetheless, the formula is: =IF(ISFORMULA(cell)," "&FORMULATEXT(cell),"") If you see: =IF(_xlfn.ISFORMULA(cell)," "&_xlfn.FORMULATEXT(cell),"") then your version of Excel does not support the Functions. www.excelfunctions.net/excel-isformula-function.html www.excelfunctions.net/excel-formulatext-function.html Best Regards.
I haven't seen a single dislike for your videos. Great work. Your videos helped a lot to me.
I am glad that the videos help!
This is an amazing channel! Amazing instructor, amazing videos.
Thanks, Sally Lee!!!! Glad it is Amazing^3!!!!!!
Your channel is incredible. Thank you so much!
You are welcome, TheWiseSophist!!! Thanks for your support with your comment, Thumbs Ups and Sub : )
@AM The page is huge, you have to scroll down to the "Basic Excel Business Analytics: Highline BI 348 Class " section.
Thanks 🙏
Hi, Thank you so much for the comprehensive training on regression techniques. I have a question. If a data set has both seasonality (quarterly sales) and trend, but the trend is expected to change after a certain period, how do I accommodate that using your techniques? The second question is if the data has more than one seasonality (more than one-quarter of low/high sales), will the technique work efficiently? Please let me know. Thanks.
You saved my life, thank you!
I am glad that the videos help!
Thanks a ton!
You are welcome a ton, Abhijeet!!! Thanks for your support with your comments, thumbs up and Sub : )
You are a saint.
Thank you Mike. Stellar job on your entire body of work.
Regarding Categorical Variables:
At 9:53 - 10:22, you state:
"We're going to have to add 1, 2, 3 extra columns to our dataset. Why 3? Because if you have whatever number of categorical variables, you have one less column than there are variables."
This seems to beget the question, why, "if you have whatever number of categorical variables" do you only need "one less column than there are variables"? What is the reason that one---in this case Q4---is not included in the same manner as the others, and why is it Q4 rather than one of the other 3?
Perhaps this is related to your statement at 11:46 that:
"The slope for quarter 4 will already be incorporated into the Y-intercept."
How so? Why Q4? Could this use further explanation? Thanks.
Best regards.
Drives me crazy
"Why 3 quarters? Because 4 - 1"
I still don’t get this. Have you understood it by now and could you please explain?
Could the creator of the video maybe please elaborate on this topic since it seems like multiple people are struggling to grasp this concept. Thank you!
You choose X to be time, however other variable could be choose too. For an example price elasticity of demand price could be use as X variable, right ?
This is great, thank you so much!
You are welcome so much!
Hi Sir, your videos r very helpful. Just a quick question, how you determine when you will use moving average and liner regression? Based on MSR? Thanks
awesome as ever.
Good video. But just want to confirm: in example 3, the p-value for Qtr3 is 0.07 which indicates insignificance of Qtr3 coefficients. That needs to be fixed right?
Simply amazing, Thank you so much
You are welcome very much!
thanks pal, awesome job
Is there a particular reason for using the Catigorical column formulas you did at about 12:35?...
=VLOOKUP(C4,$O$4:$R$7,MATCH(I$3,$O$3:$R$3,0),0)
It's an instructive formula, but wouldn't it have served equally well to simply name the Categorical Columns "Q1, Q2, & Q3" and then used
=IF(H$1=$C4,1,0)
to compare each Catigorical column header back to the Quarter in Column C?
Maybe I'm missing some additional subtle consideration here. Thanks.
Best Regards.
Great Job, thank you so much
Also, is there any big differemce between making forecast with linear regression and multiple regression?
yes, yes it is
but something tells me you don't need the answer anymore
Hi. What is the difference for calculating the slopes for each Quarter rather than getting the linear trend and averaging a percentage between the quarters from the historical data, so the four Quarter averages would be 100%.
Wouldn't it create a very similar forecast?
Hi which of your template can help me reviewing within mfg capacity[PO +/- [% FCST]] . i am trying to see if i can create order report that tied up mfg capacity , FCST, Serves level , SS ...
hello, can u make a lesson on principal component analysis or Factorial analysis?
hello, So i Do not understand. Whats is the main difference between forecasting with Linear regression and basic forecast methods ?
+Кирилл Чувычилов I would try it both ways and see which gives you the smallest Forecast Error. Regression should be best if there is a trend.
Hi, Can you tell me how did you manage to show the formula in I column and values in G column simultaneously at +3:07.
I'm looking for the same thing for very long time but no success yet.
Thanks
Satish,
You could figure this out simply by downloading the provided Excel file from the Excel Is Fun website and examining the cell contents.
Nonetheless, the formula is:
=IF(ISFORMULA(cell)," "&FORMULATEXT(cell),"")
If you see:
=IF(_xlfn.ISFORMULA(cell)," "&_xlfn.FORMULATEXT(cell),"")
then your version of Excel does not support the Functions.
www.excelfunctions.net/excel-isformula-function.html
www.excelfunctions.net/excel-formulatext-function.html
Best Regards.
The predicted values aren't reasonable. Qrt 2 - Qrt 4 are much lower than the actual sales value and that will mean a shift in the trend.
The predicted values have increase according to their respective slope lines. Qtr 2 and Qtr 4 are not lower than the previous Qtr's actual sales.