Machine learning library
Project description
machinePy 2.0
Welcome to machinePy, the worlds most user-friendly machine learning neural network library!
To learn how to use it, visit machinePy, where you'll find everything you need to start including detailed examples.
Quick rundown
foo = initnet(network_structure)
Use this function to declare your neural network structure and assign random weights and biases. It takes an array of integers. The first element determines the amount of inputs and the last element determines the amount of outputs. Every element in between represents a hidden layer with the integer value as the amount of neurons in that particular hidden layer.
foo = forward(inputs, weights)
This is the forward propagation function. It takes two arguments, an array of inputs, and an array of weights, which is the output of initnet
.
foo = train(inputs, weights, targets, learning_rate)
This is the implementation of the infamous backpropagation. It takes four arguments, an array of inputs, an array of weights, again in the form of initnet
, an array of target outputs, and an optional learning rate value, otherwise assigned 0.1.
Example
import machinepy as mp
inputs = [[0,0],[0,1],[1,0],[1,1]]
outputs = [[1],[0],[0],[1]]
weights = mp.initnet([2,3,1])
epochs = 10000
i = 0
j = 0
while(i < len(inputs)*epochs):
weights = mp.train(inputs[j],weights,outputs[j],0.1)
if(j==3):
j = 0
else:
j+=1
i+=1
i = 0
while i < len(inputs):
foo = mp.forward(inputs[i],weights)
print(foo)
i+=1
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for machinepy-2.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5ef89377ef2fec64151303c91894cf92571287d66b7c9a9c3c06d7582afda393 |
|
MD5 | bbd2a3c37ec7becfd01c315959a9b9fe |
|
BLAKE2b-256 | 9faa0506ee977509a92eabfba67d9624cfd9f11bf6a3f991fd1543541e75d830 |