From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

MACHINE LEARNING HELP!!

Hi there,

 

i wonder if anyone could point me in the right direction. I'm trying to use the MLT but I'm not quite sure how to go about getting what I want. Basically when I take in a signal I want to use the MLT to recognise the signal that's coming in and whether it is more like A or B.

 

At first i was looking into k-means and density estimation but now I am thinking kNN. But I have no idea of how to even go about this with LabVIEW. Looking at the toolkit I've also doscovered Spectral Clustering. Would this be better? iIve tried messing around with the toolkit but I'm getting no results. As I said, any help would be greatly appreciated!

 

Kind Regards,

 

KelOb

0 Kudos
Message 1 of 5
(2,971 Views)

Hi kelOb, 

 

Though is is community content, I will do my best to look for some useful resources for you. First of all, have yo u investigated the Toolkit user profile that can be found on the community page? 

With regards to the specific mathematics behind a function, I would recommend reading up on them online (for example on wikipedia) to work out which is best for your needs. If you have any other specific questions you feel we could answer, please don't hesitate to let me know. 

 

Kind regards,

Matt H
Applications Engineer
National Instruments UK
0 Kudos
Message 2 of 5
(2,642 Views)

Hi Matt,

 

Brilliant, thank you and many thanks for your reply it is greatly appreciated. I have looked at the page a few times yes and researched all methods. As far as I can see k-means and clustering the data may be the best for me. however when I use the vi's in the toolkit, even though i get a reading, it is messy. Realistically I would like a class 0 (red) to represent a 'no event' in my sensor and a class 1 (green) to represent an event. and to cluster the events depending on strngth (they are seismic signals). I have attached a screenshot of the XY graph I have so far. 

 

Any help at all is appreciated as I can't see where I'm going wrong.

 

Kindest Regards,

 

Kelly

0 Kudos
Message 3 of 5
(2,631 Views)

Hi Kelly, 

 

One alternative to this particular piece of community content could be this library intended for clustering. Otherwise the first post on this page shows some alternative free sofrware packages you could use to obtain such functionality. Let me know if these are of any use to you, in the mean time I will look for further examples. 

 

Kind regards,

Matt H
Applications Engineer
National Instruments UK
0 Kudos
Message 4 of 5
(2,589 Views)

Hi Matt,

 

That's great! Many thanks, I shall look into that now.

Kindest regards.


kelly

0 Kudos
Message 5 of 5
(2,581 Views)