Market Analysis Report for National Clothing Chain
Udacity Project
Github Repo
Linear regression
Linear Regression method helps us to find whether the relationship among X variable(Independent) & Y variable(Dependent) is exists & level of effect.
here
- Postive relationship (X increase leads to Y increase)
- Negative relationnship (X increase leads to Y decrease)
eg: customers income level vs purchases of products
- R = correlation Cofficient(ranges from -1 to 1),
- R^2 = Coefficient of Determination(ranges from 0 to 1)
correlation b/w avg. income by state & last 6 months avg. purchase by customers
*Coefficient of Determination = 0.78**
Using Quick measure
Under Mathematical operations - correlation Cofficient(R)
then applying square to R
that is R^2 = Coefficient of Determination
Now I will show you equation which is helpful to predict the future sales by customer income
using
y = mx + b
y = mx +b
n = number of rows
x = the variable x you input
m = [n x (sum of xy) - (sum of x) * (sum of y)] / [ n * (sum of x^2) - (sum of x)^2]
b = [ (sum of y) * (sum of x^2) - (sum of x) (sum of xy)] / [ n * (sum of x^2) - (sum of x)^2]
m = 0.010726
b = 722.14
Now using the equation we can predict the purchases from customers depends on there income level.
eg: if the Avg. income level(X) = 150k
then avg. purchases(y) we predict = 0.010726(150k) + (-722.14) = 886.8 y = mx + b
customer Income Bin
DAX
Customer Income Bin =
IF([Predicated Income X]<50000,0,
IF([Predicated Income X]>=50000 && [Predicated Income X]<100000,50000,
IF([Predicated Income X]>=100000 && [Predicated Income X]<150000,100000,
IF([Predicated Income X]>=150000 && [Predicated Income X]<200000,150000,
IF([Predicated Income X]>=200000 && [Predicated Income X]<250000,200000,
IF([Predicated Income X]>=250000 && [Predicated Income X]<300000,250000,
IF([Predicated Income X]>=300000 && [Predicated Income X]<350000,300000,
IF([Predicated Income X]>=350000 && [Predicated Income X]<400000,350000,
IF([Predicated Income X]>40000,400000,0
)))))))))
we can use SWITCH function
customer raing vs returns
*Coefficient of Determination = 0.69**