Improving R2 score in Regression model

raghu

New Member
#1
Hello,
I got a case where we need to do regression on previous energy consumption data and we predict our future consumption.
Overview:
-What we do is, We develop some energy saving products and prove that by using those products, we reduced the Energy consumption(carbon emissions) for the upcoming years and we claim certificates (VEECs).
-For this two choose two periods baseline and Operating. Baseline is the period where our products are not installed and operating is the period where our products are installed and start minimizing consumption.
-We collect energy consumption data for both these periods. Also weather is another factor which plays a key role in modelling so we collect BOM(weather data) for both the periods.
-We develop a regression model that predicts our future Consumption. Two models are developed for both data baseline and operating and finally we adjust the baseline model to operating model conditions.
-Dependent variable (Y) is energy consumption and independent variables(X1,X2) are CDD,HDD(weather data).

File:
This forum not allowing me to upload the excel files. You can download the file through the following link:
https://we.tl/t-V80e8u3Cqy

Queries:
1)The R2 and Adj R2 values for the baseline model are poor(055,0.45) I need atleast 0.75 to pass. So how can improve that?
2)Is my model good enough even with low R2 values? Actually the body where we need to submit demands that it should be more than 0.75
3) Is it always possible to get R2 values greater than 0.75?
4) Can we make submission with low R2 values? If so, what reasonable justification can be given?

Request you to kindly help me with these queries..


Thank you all..
 
Last edited: