Statistical Analysis of lots of Parameters

Hello everyone,

I have just started a project looking at some changes in data for certain parameters on some machinary and I want to find out if any other parameters are also changing/contributing to this change. I get 1 scan of each parameter each time the machine is run, there is around 25 parameters and I have lots of different machine data to work with.

As this is for work unfortunately i am not allowed to share the data.

What is the best way to approach this using MATLAB?

Thank you in advance.

Answer 1

[Udit06]

Hi Huw,

As per my understanding, you want to analyze multiple parameters in your data statistically. You can consider following the steps below for the same:

1) Preprocess your data by cleaning the outlier, handling missing values and normalizing the data.

2) Once the preprocessing is done, you can perform an exploratory data analysis on the data by using functions like "min", "mean", "var" to get an understanding of data distribution. You can additionally plot your data to visualize any patterns or correlation.

3) You can generate the correlation matrix using "corrcoef" function to identify potentially related parameters. You can refer to the following documentation for details related to "corrcoef" function.

https://www.mathworks.com/help/matlab/ref/corrcoef.html

4) You can use statistical tests like "ttest" to determine if changes in parameters are statistically significant. Please refer to the following documentation for more details:

https://www.mathworks.com/help/stats/hypothesis-testing.html

I hope this helps.