NEAI only 50% percent accuracy on Validation

[MHantke_001]

Hi all,

i have a small question about Validation of a trained model.

Atm i only work on PC with NEAI Studio Version 5.1.1.

I used only data out of a *.csv file

I trained (benchmark) the model with around 170 Lines and 16 axis to get an accuracy of ~97%.

I then validated the library with some test data with same axis count and data preparation. 

The learning file has the minimum amount of lines as suggested from the benchmark, i tried more lines with same result.

The validation shows that even good data is treated as abnomaly so i only get an accuracy of max. 50%.

Are there any hints what i am doing wrong?

 

Many thanks in advance for answers.

BR’s

Martin

[Julian E.]

Hi ​@MHantke_001,

 

This is a common issue in machine learning.

What could happens is what we call overfitting.

 

Because you don’t have much data, the algorithms learns your data more than how to solve the issue. So when you test with new data, it performs poorly.

 

I would suggest looking at this document I wrote to find some ways to improve your results.

AI:Datalogging guidelines for a successful NanoEdge AI project - stm32mcu

 

Have a good day,

Julian

[MHantke_001]

Hi Julian,

many thanks for your answer.

I apologize for the late response i was out of office.

 

We are testing and working with NanoEdgeAI studio now for some time and it works better and better with your guideline.

BR’s

Martin