Just a simple neural network (single-layer perceptron) in Rust for learning purposes.
Build the project with cargo build and run it with cargo run.
Plotting the error rate over the epochs:
The output in the console:
Input: [0.9, 0.1], Prediction: 0.9
Input: [0.5, 0.5], Prediction: 0.9
Input: [0.2, 0.3], Prediction: 0.5
Input: [0.3, 0.6], Prediction: 0.9
Input: [0.1, 0.7], Prediction: 0.8
Input: [0.3, 0.1], Prediction: 0.4
Input: [0.1, 0.5], Prediction: 0.6
Input: [0.9, 0.0], Prediction: 0.9
Input: [0.3, 0.3], Prediction: 0.6
Input: [0.0, 0.1], Prediction: 0.1
Input: [0.1, 0.2], Prediction: 0.3
Input: [0.2, 0.0], Prediction: 0.2
Input: [0.6, 0.1], Prediction: 0.7
Input: [0.5, 0.3], Prediction: 0.8
Input: [0.9, 0.1], Prediction: 0.9
Input: [0.1, 0.4], Prediction: 0.5
Input: [0.2, 0.4], Prediction: 0.6
Input: [0.7, 0.0], Prediction: 0.7
Input: [0.6, 0.3], Prediction: 0.9
Input: [0.2, 0.2], Prediction: 0.4
Input: [0.1, 0.0], Prediction: 0.1
Input: [0.2, 0.6], Prediction: 0.8
Input: [0.5, 0.0], Prediction: 0.5
Input: [0.6, 0.4], Prediction: 0.9
Input: [0.4, 0.5], Prediction: 0.9MIT License.