Machine Learning Example (Fruit)

[Rod]

I have in mind a little project to create a ML example. An example of a neural net (small) that is fed data about fruit. It learns to distinguish and identify fruit.

A small set of data, weight color, texture, size, shape etc accompanied by a set of images of the actual fruit. The perceptron will handle the data the images are just for fun. It will look like the computer is looking at the images but it is really reading the data.

It will also really self learn, it will train and it will then look at a second set of data and tell you what the fruit is.

So any thoughts, any pointers any code to help out?

[plus]

Very ambitious project!

With perceptron beginner's exercise, recognizing characters handwritten, it was difficult to align the image with Perceptron's sensors making sure scale is right and center of image match.

[Rod]

Yes, interpreting real images would be hard. I plan to bypass all that and just deliver the data for each image ready packaged. I should get some data together and repost.

[plus]

Well I hope this project bears fruit.