This is a demo of some neural network and machine learning widgets. Tyler Truong created this to show neural networks and machine learning in 2d plots and with languages.
The 2d plots was a college question on a test where I bubbled neural network instead of support vector machine because I didn't read the book at home. I learned in this demo that my professor was correct and a voronoi tesselation based support vector machine works best for 2d plots. It's able to even support vectorize random points.
The language page was a college class in advanced statistics based on statistical regression and classification. I aced the class by showing the same methods of logistics regression and co-dependencies can be used to detect the language or person who wrote a text message. In class I used Trump and Hillary to detect ‘Build the wall” and ‘African American” to see which candidate was more likely to say those phrases. That's a silly machine learning language classifier so in this example, I used the more appropriate English, or Spanish, or French, or Vietnamese classifiers. This demo improves on the college example by detecting 4 possible outputs instead of just 2 outputs.
The source code for this example can be found on my Github page.