Hi Vishal,  assuming your sensing dataset has a large number of variables, also called data points (i.e the column objects). Your rows are called the observations. PCA helps you to reduce the very large data point space into low dimension space. In this low space, important components are compressed to reduced computational speed, space and resources. You will prepare your sensing datasets in rows and columns, as I have said, put your variables in columns and your observations in rows and run it on a convinient platform such as R, Matlab, Weka, Azure Machine Learning Studio etc.

Best!

Hi Vishal,

Tricky question. In general, People use PCA blindly without knowing what exactly it does.

I would say it more or less depend on the objective.

Although your question is confusing (because on one hand, you are asking how to apply PCA and another hand you are asking how to use sensing data) but I understand that you have sensing data and you want to reduce the variables so that using PCA maximum variability of the data can be explained.

If it is the case, I can provide you a simple and short MATLAB script.

Cheers!!

Thanks Ankit

I have recorded sensing response of various fabricated sensors for different gases, and now i want to extract an information from it to relate it with pattern recogization. Most of the  research articles use PCA analysis for the same. I also want to understand the importance of PCA analyis in analyzing sensing data.

i'm interesting and following the answers