DOE-3: Design of Experiments: Coded and Uncoded values & establishing regression equation

preview_player
Показать описание
I am happy to share my third video on Design of Experiments (DOE-3). This is the third video in our series on Design of Experiments. It is recommended that viewers should watch the first two videos (DOE-1: DOE Introduction and DOE-2: DOE Case Study of Spot Welding) before watching this video for better learning and understanding. In this video, we have explained how to convert uncoded values of factor-levels to coded values and vice versa. We have also explained how to create a linear regression equation from the results of a full factorial two-level experiment. All videos on Design of Experiments are available in a playlist with the same name.

There is a small correction at 9:35: The equation is incorrectly shown as (3Xuc-13)/3. It should be same as (Xuc-13)/3. We apologise for the error.
Рекомендации по теме
Комментарии
Автор

Hemant sir, you are a brilliant teacher. Very grateful to you. Huge respect.

MahadevHarHar
Автор

As easy an explanation can get :). Very informative for both green belt and black belt aspirants.

ankitverma
Автор

Nice video. It gave me clarity on coded and uncoded equations

AlokKumar-jehd
Автор

Great video. However, I'm confused at the 9:34 mark with the conversion of coded to uncoded values in the Red Box. There seems to be an error in the 2nd and 3rd line items? Shouldn't 29.25 x2C = 29.25(x2UC -3)? Seems like it is missing the 29.25. Also, 58.5 x3C should = 58.5(x3UC-10). Or am I mistaken?

jerryok
Автор

Hello,
Thanks for your very helpful video
I have created a full factorial design of 3 levels-3 factors, yet my values do not vary with the same interval, i.e. for a given factor x, the three levels x1, x2, x3 do not verify the expression: x2-x1=x3-x2. How can I change my model from coded values to real values?

firstnameylastnameo
Автор

Hi sir, thanj you for your presentation, i want to know how we can find the slop of Beta 1*Beta 2 x1*x2

karimkerrouche
Автор

Can you please explain what would happen to Bo if you reduced the model (for example remove a term) - does Bo change? Thank you!

dianas
Автор

Hi sir, when we want to predict a new set of values, which regression equation should we use? (coded or uncoded)

nmana
Автор

When do you use -1/1 coding as opposed to 0/1? I've always used 0/1 and I am seeing more and more people use -1/1

fhajjat
Автор

Hi, i have one more question, when transforming coded to uncoded equation, should 'y' term also be transformed?

nmana
Автор

Hi. How do we optimise the parameters instead of using a high or a low setting

teelit
Автор

Hello sir video 06:10

Beta23.X1.X3.

Is actually

Beta23X2X3

M i right?

lokeshgupta
Автор

Sir the equation for uncoded to coded conversion is bit confusing.

Pl see the current coding conversion equation at video time 03:35 it is

Current (coded)=(current uncoded-13)/3

this is clear.


But at video time 09:35 it is X1c =(3Xuc-13)/3.


So why Xuc is additionally multiplied by 3 here, could u pl explain.

lokeshgupta