Hello there, Scratch. This is my second project - A little bit smarter Neural Network. Click to go to the next slide. Press X to skip explanation and see the Neural Network. Press A to try it out yourself.
This project was also rewritten from python. A little bit of debugging :) Font: Anurati This project aims to use another activation function to fix sigmoids disadvantages. As you might have seen last time - >no negative values >no correct output for 0 0 0 Now fixed by tanh! >Now also checks for errors. >Added graph for tanh and its derivative >Added mobile support Also see: Part1: https://scratch.mit.edu/projects/374597163/ Thanks to me for debugging and coding. #artificialintelligence #ai #neuralnetwork
