https://github.com/SC4RECOIN/NeuroEvolution-BTC-Trader

Building a population of models that trade crypto and mutate iteratively
https://github.com/SC4RECOIN/NeuroEvolution-BTC-Trader

algorithmic-trading algotrading bitcoin bot genetic genetic-algorithm neuroevolution reinforcement-learning trading trading-bot

Last synced: about 2 months ago
JSON representation

Building a population of models that trade crypto and mutate iteratively

• Host: GitHub
• URL: https://github.com/SC4RECOIN/NeuroEvolution-BTC-Trader
• Owner: SC4RECOIN
• Created: 2018-06-10T02:55:49.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2021-03-09T21:16:03.000Z (almost 4 years ago)
• Last Synced: 2024-04-18T04:17:15.419Z (8 months ago)
• Topics: algorithmic-trading, algotrading, bitcoin, bot, genetic, genetic-algorithm, neuroevolution, reinforcement-learning, trading, trading-bot
• Language: Go
• Homepage:
• Size: 101 MB
• Stars: 155
• Watchers: 14
• Forks: 44
• Open Issues: 4
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# NeuroEvolution BTC Trader

The goal of this project was to create a model that would trade on technicals through neuroevolution. The initial population is built given a specified network topology and assigned random weights. For each generation, these weights are randomly mutated and each network is assigned a fitness value based on its performance. A pooling algorithm is then used to select models for the next generation and the process is repeated. Models with a higher fitness tend to make it to the next generation where their weights are mutated and ideally create a higher performing generation.

See [python branch](https://github.com/SC4RECOIN/NeuroEvolution-BTC-Trader/tree/python) for original implementation