# Online Class Exercise Chapter 3 Module 1

Okay! So, we’re gonna do the analytics of
chapter 3 which is forecasting. We’re gonna do various techniques here. The
first method that we’re going to do is simple moving average. Simple moving
average basically looks at the past three months and then taking the average
of the past three months if it is a three week moving average or we’ll take
the past nine months are nine weeks if it is going to be the nine week moving
average. Here we have weeks so we’re going to be looking at weeks…the period
is week… the period can be a week or a day or a month or a quarter or a year
whatever you get so whatever data that you have. So,here I have data
for weekly data so my first week’s demand was 800; my second week’s demand
was 1400; my fifteenth week demand is 2,000. I’m going to forecast the three
week moving average first and then I’m going to forecast the nine week moving
average after that so if I want to forecast the three week moving average
for period four I’m going to take the average function the average function of
the past three which is going to be the 800, fourteen hundred and one thousand going to start with C5 through C7 I close the bracket
and that is going to be 1067 now I can pretty much copy and paste it and I
should be able to get the moving average or the three week moving average for
various weeks. So for week five the demand was 1,500
the three week moving average was 1,300 see the difference! There is quite a bit
of difference. Similarly if you go to this end if you
have the demand was 2,000 but the three week moving average was
2033 it’s not better than the first 3week average simply because we are going through this particular demand has got tighter demand a closer
demand that’s a reason if the demand is high it’s gonna be high so it’s not inaccurate it is not very accurate at all. Let’s take the nine week
moving average is going to be again that I’m gonna take the average function of
the past nine weeks so past nine weeks is one to nine okay so I’m going to take
the past nine weeks which is one to nine and if I take that the it is going to be
1367 1367 is the nine week moving average for period week ten so similarly
I can do copy and pick paste it and I can see that for the nine week moving
average for week 15 is 1733 while the three week is 2033 we really don’t know
whether which is accurate here because even though I’ve taken the nine months
average we don’t know it could include the seasonality trend randomness
whatever it is so we really don’t know it’s accurate. When we talk
about simple moving average now. Let’s look at the next technique which is
weighted moving average the weighted moving average basically has got let’s
say form this is a month this was in weeks
this is in months okay so the first month demand was 100 the second month
demand was 90 the third month demand was 105 and the fourth month demand
was 95 we need to find the fifth month forecast that’s what we are trying to do
now I also have as an input these things here what I have is that I’m gonna take
10% weightage of month one’s demand 20% weightage of months two demand 30%
weightage of month three demand and 40% weightage of month four. I
need to consider the weight here and then I’m going to do the moving average,
you have guessed it now, so we are going to do an average but the average is
basically going to be based on all the different weights and all the different
demands that we have so as we find the weighted moving average for the fifth
month so the equation that we are going to put right now is going to be for the
demand or the forecasted demand for the fifth month is going to be equal to I’m
going to take the 95 which is the fourth week’s demand times the 40% okay because
it’s a 40% of the month four plus I’m going to take the 105 times
the 30% which is for the month three plus and I’m going to take 90 times 20%
which is the month two plus the 100 times the 10% which is going to be for
the month through the first month okay if I take that the moving the
weighted moving average for the fifth month is ninety seven point five. So you
can see that if this is little more accurate simply because I can change the
weight. So, if I want to I can change this to
60% and then make this to 10% because see even if I do that it is gonna come
as 96.5 because I’m pretty much I’m putting the weights differently now it
has become 95.5 now the 96.5 right so I can go back to whatever I have which is
the 40% and you can see the difference it is ninety seven point five to ninety
six point five based on the weight age so if I give a lot of weights for the
first month then it is going to change drastically also you can see that you so
you can pretty much play with it. You want to do this when you’re going to do the
Excel after you go through this tutorial. Do the excel sheet
and then you can play by changing the weight here and you can see how it
all changes okay so that is basically the weighted moving average. We are now
going to move on to simple exponential smoothing the simple exponential
smoothing again here we have week 1 to 9 and the week the demands for the week 1
to 9 is given to you we are also given one more thing which is the smoothing
factor and so we need to calculate the smooth forecast using the smoothing
factor how do we do that so we have an equation
that you need to follow the equation is f of T which is for a month.
Let’s say for month two, you’re going to take F of P minus 1 which is the the forecast of one month
prior which is the first month so if you want to find the forecast for month two you’re going to take the forecast of the past month which is t minus 1 which is
the first week in this case. These are weeks… it’s not month… so let’s go back
again so if you’re going to look at the the forecast
– then you’re going to take the forecast of the prior week which is the first
week plus the smoothing factor times the actual demand of the prior week – the
forecast of the prior week so I’m going to use the forecast of prior week here
you’re going to use the forecast of prior week here you’re going to do the
actual of the prior week here and then you’re going to use the smoothing factor
to find the weekly forecast. For instance, for week one, we do not have
the anything before the first week and so I’m going to
take that as equal to 820 so it is the same as what my demand is because I do
not know the prior weeks so I’m going to use that as a assumption and if the
actual demand is is you know basically the actual demand we don’t know if the
actual demand was 820 or 811 we don’t know. That’s the reason I’m
assuming that is going to be 820 so first week we’re going to take it as the
same as a demand the second week we’re going to use this formula what is a
formula here the formula here is going to be the smooth forecast is going to be
equal to the forecast of the prior week which is 820 plus the factor,
which is a smoothing factor which is the Alpha, times the actual of the last week is 820 which is actual demand – the forecast of the last week which is
going to be 820 I’m going to close all the brackets and that is going to be
820. Now, if I copy and paste I’m going to get these numbers. This is
where the problem is! So, I cannot just copy and paste simply because… see
this one here you need to use what is called as absolute reference. We’re
going to use an absolute reference if you just copy and paste you’re gonna get
the same value or some other value because when you copy and paste it is
going to change okay because point two will become point zero… it
changes… it changes in every place… this alpha is a constant so we need to
make sure that this alpha which is which is actually P3 which is here P3 should
be an absolute value. So, I’m going to make it as an absolute which is dollar P dollar
three and now if I go and copy that. I should be fine now!
So, make sure when you’re doing copying and pasting you need to make sure how
you’re going to use the reference the reference should be pretty well defined
and and if you want to be an absolute if you see anything constant like
this you need to use an absolute reference. If you do not
have a constant, of course in Excel you just copy and paste. Now, everything will
change okay so now we know the answer is for the tenth week it is seven hundred
and fifty nine point six one seven hundred and fifty nine point six one
that is your and for the third week it is eight hundred eleven and so on and so
forth. This is how you calculate the simple exponential smoothing using the
smoothing factor. I’m going to move on.. this is the first module…I’m going to go to the module two and then we will we are going to do the
linear seasonality in the next
video 