This project demonstrates a simple feedforward neural network trained using backpropagation. Two examples are provided:
bits.py: Converts binary input to its corresponding decimal representation.successive.py: Predicts the next number in a sequence.
This neural network takes 2 bits as input and gives the output converted in decimal.
For example:
input (0b10): [1, 0] output (2): [0, 0, 1, 0]
| Input Neurons | Hidden Layer 1 | Output Neurons |
|---|---|---|
| 2 | 4 | 4 |
This neural network takes a number as input and returns the successive number as output. For example, with four numbers:
input (number 2): [0, 1, 0, 0] output (number 3): [0, 0, 1, 0]
| Input Neurons | Hidden Layer 1 | Output Neurons |
|---|---|---|
| 6 | 6 | 6 |
I wrote this implementation with a focus on readability and functionality, not optimization :)
I wanted to list some of the youtube videos that helped me understand neural networks & backpropagation
- 3Brown1Blue's beautiful playlist ❤️
- Mahesh Huddar's blackpropagation solved example