AI-Studi_BME688_Classification

[MaLeTa]

Hello,

I am using AI Studio to generate different samples from my measurement series. Now I have seen only every eighth cycle in my sample data. For this reason I looked in the AI manual and in the example.specimen-data also not all Cycle ID's are included. What is the reason for this behavior? In my data no cycles were omitted, the corresponding value in AI-Studio is 0%.

The attached picture is from the AI manual. Where are the other cycles between the listed cycles? I can imagine that this behavior could have a negative impact on AI training, since the amount of data is smaller than expected. What are your thoughts on this? How can I take an influence of the listed cycles in one copy?

With kind regards

MaLeTa

[BSTRobin]

Hi MaLeTa

If the recorded sample data meets the format requirements and can be successfully imported into BME AI-Studio, it will be able to train normally in BME AI-Studio without any impact.

[MaLeTa]

Good morning,

I know that the data are usable. But the question is: Why there are not all cycles included in the data? Why only every fourth or eight cycle is used by the AI-Studio to train the algorithm?

[BSTRobin · Accepted solution]

Hi MaLeTa,

If you use the official BME688 dev kit to record data, it can be imported into BME AI-Studio normally. As long as the data can be imported into BME AI-Studio, you can continue training the data without worrying about how the AI algorithm handles them, as its details are not publicly disclosed.