Forum Replies Created
-
Posts
-
Hi John,
Thank you for you ideas and feedback.
Confidence level is now available is SCiO Lab Mobile when you test your model, you will see the model’s response and a number in brackets. That number is the confidence level of the answer – 0 means no confidence (theoretically should never happen), and 100 means full confidence.
In addition to the confidence level, you can train the model to disregard scans that look significantly different than the scans used to train the model (the response will be “null” in that case).
To activate the “outlier detection” feature, use the expert mode, click on the settings button (cog wheel on the right), and check the “outlier detection” option.
About sharing and collaboration – we are very much in favor of creating it in the future, at this point we can help by copying and merging collections from different users. Please contact us at dev@consumerphysics.com if you have a specific request in mind (some developers already have shared the collections with each other).
Hi Paul (and Roger)
Roger is right – method 2 is the right way to create your data collection. SCiO’s algorithmic engine requires DIFFERENT samples to have a successful model. Scanning all notes under the same sample id will result in a poor model. To write it differently – 100 scans of a single sample is not the same as 10 scans of 10 samples.
Using PTFE as a reflective surface is a good idea – it does not have prominent feature in NIR and is supposed to be very reflective.
I would also scan notes of a different denomination to see if there’s a difference between them, and obviously, if you have access to counterfeited notes, it will be very good to scan them as well.
Keep us posted
Hi everyone,
We recognize the need for this feature and are working on it on “full steam”. I hope we will be able to release this feature soon.
Your questions and feature requests are making a difference when we prioritize our tasks, thank you for allowing us to make SCiO better!
Guy
Hi,
I’m Guy , product manager of SCiO SDK.
Here are the answers:
- X is the wavelength represented in nm (nanometers).
- Y is dependent on the state of the toggle buttons in SCiO Lab.
When:
- No buttons selected: Raw reflectance spectrum
- Processed (only): Assumes Beer-Lambert model is valid, and transforms the measured signal to be linear with concentration by doing a log transform and adjusting the result for noise and deviations from the model. You can learn more about Beer-Lambert here: http://www.chemguide.co.uk/analysis/uvvisible/beerlambert.html
- Normalized (only): Performs normalization of the signal. This is meant to compensate for changing measurement conditions (e.g. varied scanning distances) that typically occur from sample to sample. Y axis still means reflectance but in normalized units instead of raw reflectance.
- Both Processed and Normalized: First assumes Beer-Lambert model (like processed) and then normalizes the results to compensate for differences in the optical path between samples. This is useful, for example, when there is variation in the thickness of the samples.
Typically, different models and types of samples will require different pre-processing methods. You should both choose the pre-processing method to match your experimental setup and optimize the performance of your model. If you planned and gathered your data correctly, these efforts will coincide.
