Neural net - putting weights into strings

[MarkBai]

Hello, this is the VI that I am trying to make.

Im trying to make a VI that whenever I input a word it would display a corresponding letter. I know it sounds easy but the catch is I have to use neural net and training in this VI. I know I have to put weights to each word and train the system but how can I put weights into the array of strings when the weights are numeric? PLS HELP 😞

[danil33]

where is the VI that you have tried?

[Ray.R]

You could use an array of clusters.  The cluster would contain 2 items:  a string control & a numeric control.

The string control would hold the word, while the numeric control would hold the weigth.

Since they are part of the same array element, it is easier to index.

[MarkBai]

Ray.R

how is that exactly? Do i put a cluster inside an array and put an array string and numeric array together?

[Ray.R]

I will post a snippet

If you build it in the block diagram, start with a cluster constant, then place a string control and a numeric control inside the cluster.

Place an array constant on the block diagram and drag the cluster inside the array.  You can then change it into a control.

[MarkBai]

danil33

I'm only beginning to do it so I cant provide one right now. But what I'm trying to do is similar to this VI except that the inputs and outputs are strings and the weights are numeric

[Ray.R]

Thanks Mark for the complete example..

I can skip the snippet.  😉

Let me guess, this must be an assignment, right?

[MarkBai]

yeah -_- im new to labview and only know the basics so I dont have much information yet to actually do this by myself 😞

[PhillipBrooks]

If the list of words will be large and performance important then you might want to consider using variant attributes as a way to store key/value pairs.

Your words would be the key and the value would the weight.

• Implementing sets with variant attributes
• Darren's Weekly Nugget 10/09/2006
• MAP with LabVIEW

[MarkBai]

PhilipBrooks,

by large do you mean the number of words i'll input? or the length of the words? its only gonna ba a word for word translation. (e.g Alpha --> A)

although yes I was assigned to do this with at least 100 words  -_-