Creation of SB-Neuron. Ours. Branded.(v2)

[litdev]

OK, Xs and Os in a grid is a good idea - what size of grid and how many samples - I could create one traing set and you another, then when we each have a trained AI we could swap training sets and see how we do using each other's data?

[AbsoluteBeginner]

OK, Xs and Os in a grid is a good idea - what size of grid and how many samples...?

Yeah. 

Does this mean that you will give me the opportunity to test our SB-neurons using tests we can trust?
This is not only interesting, but also important to me.

Naturally, I won't be able to do the work quickly. The shelling of our city is intensifying.
But, I still hope that I won't have to leave my home and become a homeless refugee.
I hope that I will be lucky enough to finish creating unusual SB-neurons and get great creative pleasure.

I am ready to perform any task you want.
The only thing I want is to compete with a network created using the "SmallBasicANN" extension or simply in Small Basic.

So, I'm waiting for a description of the terms of your task. 

[z-s]

Salute to AB for his courage

[litdev]

I am very happy for you to set the spec for the X O detection.  You can decide the grid size in pixels, and how many samples we will create for each other to test with - this is so that neither of us can do the training with the data that will be used for the test.

So for example we could decide on a 10x10 grid, train out AI in any way we like, and each generate 20 test cases which we used at least as part of our training and our AIs can handle.  At this point we can swap our 20 test cases and see what happens.

It will be tricky to create the training and test cases if not made by hand - part of the fun in this will be creating the training set - I would probably write a program to do it - maybe from lots of different fonts positioned and scaled randomly and maybe some hand drawn and mayer other methods...

[AbsoluteBeginner]

Thank you very much for the interesting task. 

First, I will try to create a software module that will allow us to easily draw any "cross" and any "circle" on a 10x10 pixel matrix. 
Then I will add a function to save the drawn symbol to the training set.

Since the training of the SB-neural network is very fast, I want to visualize the training process.

At this point, my plan ends.
I will most likely figure out my next steps while I am implementing the first two steps of my plan. 

[AbsoluteBeginner]

Overnight, in my head, from ideas to hold a competition of neurons, an amazing, grandiose entertainment and educational project developed. 

I want to create a program that allows the user to train a neural network and observe the process of this training.

I want the user to be able to train the neural network to recognize individual elements (individual points, lines, ends of lines, corners of shapes, circles, etc.).
I want the user to be able to command the neural network to learn to recognize what shape is depicted on the matrix using the FEATURES of such a shape.

In general, I want to create a "tool" for studying and testing a neural network that will be defined in the program.
I will try to make it so that the user can easily insert the code of his own neural network into the program code.
- - - - -

I understand that this is a lot of work. But, I plan to get a lot of help from Gemini or ChatGPT.

I will try to show you the different stages of development of this project in a timely manner. 

[AbsoluteBeginner]

Hi all. 

Today I present to you the first intermediate version of the NeuroLab program code.

DWXL365.000

The code was written by the Gemini neural network. Of course, I had to make some corrections to this code. But, for the sake of interest, I try to make a minimum number of corrections.

At the moment, the program allows you to create images on the matrix by changing the color of a pixel with a mouse click.
The pixel fill mode when clicked is set by a switch on the toolbar.

I will be grateful for your advice, comments, suggestions and wishes. 

Good luck to you all.

[litdev]

AB,

This is a great start - perhaps a mode where select pixels as the mouse moves, like a draw motion.

[AbsoluteBeginner]

AB,

This is a great start - perhaps a mode where select pixels as the mouse moves, like a draw motion.

Yes, I also want to see if a neural network can handle this task. 

If the neural network can't handle it, then I'll write this function manually. 

[AbsoluteBeginner]

When trying to draw with the mouse button pressed, there are problems with synchronization. 

What requirements must be met when organizing the simultaneous operation of an event handler and the process of drawing an image on the Graphics window?