khuangaf/CryptocurrencyPrediction
Predict Cryptocurrency Price with Deep Learning
Implements CNN, LSTM, and GRU architectures in Keras/TensorFlow to forecast Bitcoin prices 80 minutes ahead using 1280-minute historical windows sampled at 5-minute intervals from Poloniex. Features comparative benchmarking across model depths and activation functions (ReLU, Leaky ReLU, tanh), with data preprocessing via MinMaxScaler and MSE loss optimization. LSTM variants achieved the lowest test loss (~15k), outperforming baseline lag and linear regression approaches.
1,028 stars. No commits in the last 6 months.
Stars
1,028
Forks
363
Language
Jupyter Notebook
License
MIT
Category
Last pushed
Mar 10, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/khuangaf/CryptocurrencyPrediction"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
shobrook/BitVision
Trade Bitcoin and run forecasting models from the terminal
philipperemy/deep-learning-bitcoin
Exploiting Bitcoin prices patterns with Deep Learning.
emirhanai/Cryptocurrency-Prediction-with-Artificial-Intelligence
First Version.. Cryptocurrency Prediction with Artificial Intelligence (Deep Learning via LSTM...
Omar-Karimov/ChartScanAI
ChartScanAI is an advanced app for detecting patterns in stock and cryptocurrency charts using...
bukosabino/btctrading
Time Series Forecast with Bitcoin value, to detect upward/down trends with Machine Learning Algorithms