Python Random Brain Module
Project description
Random-brain
About
Random brain is the neural network implementation of a random forest. Its purpose is to combine the strengths of multiple nerual networks.
Background on random forests
A random forest is a machine learning model that is composed of multiple decision trees. These trees in the forest all predict an outcome and the majority rules.
Similarities
Just as the random forest is a vote based ML algorithm, the random brain is a vote based algorithm as well, but uses neural networks specified by the user rather than decision forests.
Setting up Random brain
pip install random-brain
API
Init the brain module and class.
from random_brain import random_brain
brain = random_brain.random_brain()
import models()
Import models will take in a directory or a single .h5 file. Sub directories will be ignored.
brain.import_models(model_path = 'path/to/model.h5')
brain.import_models(model_path = 'path/to/directory')
show_brain()
Shows the keys used in the brain. This should just be the name of each imported model
brain.show_brain()
clear_brain()
Clear a single model or more by entering in the model name as a list. Leave blank to clear all models.
brain.clear_brain(item_list = ['model to remove'])
vote
Add in yTest to cast votes. Vote() will only return the votes as a numpy array and not actual predictions. This is useful if you want to run your own statistics on the votes.
brain.vote(yTest = [1, 2, 3, 4, ...])
predict (in development)
Add in your yTest to make predictions. This will attempt to make a prediction based off of the networks and will return a single answer. This is still in development.
In the future prediction and threading options will be added and improved.
brain.predict(yTest = [1, 2, 3, 4, ...])
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 random_brain-0.1.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 958c011acf902f1e308dc31293a41d0efcbb6512684b5d5d5b8897e481449276 |
|
MD5 | 5f65f81fd60c439f933f4d56bdcbb898 |
|
BLAKE2b-256 | 23eb4f0d6a7a7e24fee938ffd9b9aafbf8fbaffcb220ef74fa1125aaa80d1977 |