Sign in Subscribe

Methodology Behind the Price prediction

Methodology Behind the Price prediction

Notoko's peekin' into the Crypto future is like mixin' a secret sauce! 😜 We're throwin' in the 2-year treasury yield (like a Fed whisperer 🤫), spicin' it up with historical Bitcoin prices (like a treasure map 🗺️), and stirrin' in the crypto fear index (like a mood ring for the market 😂).

Now, we're lettin' our fancy AI models (LSTM & GRU) do their magic!

LSTM (Long Short-Term Memory) and GRU (Gated Recurrent Unit) are the cool kids in the neural network world. Think of them as your ultra-smart friends who never forget important stuff. They can look at past BTC price trends, filter out the noise, and focus on the juicy patterns that actually matter. 📊

LSTM's like that one friend who remembers every detail, even from way back when. Perfect for trackin' those long-term trends in Bitcoin's price history! 📈 Meanwhile, GRU's the quick-witted one, always ready to adapt to new info. Ideal for catchin' those sudden market mood swings! 🎢

Why these models? Well, crypto prices are as unpredictable as the weather 🌦️, and LSTM and GRU are like seasoned meteorologists. They’re built to handle the ups and downs, the wild swings, and the “what just happened?!” moments in the market. By keeping track of historical data and making sense of it, they help us make educated guesses about where BTC is headed next. 🚀 So, when it comes to forecasting BTC's future, LSTM and GRU are your go-to AI sidekicks!

so Let's look at the fine variables of one BTC prediction run :

Epoch 1/300

4/4 [==============================] - 5s 411ms/step - loss: 0.8048 - val_loss: 0.0557 - lr: 0.0010
Epoch 2/300
4/4 [==============================] - 0s 42ms/step - loss: 0.0861 - val_loss: 0.1466 - lr: 0.0010
Epoch 3/300
.
.
.
.
Epoch 24/300
4/4 [==============================] - ETA: 0s - loss: 0.0168
Epoch 24: ReduceLROnPlateau reducing learning rate to 0.0002500000118743628.
4/4 [==============================] - 0s 44ms/step - loss: 0.0168 - val_loss: 0.0201 - lr: 5.0000e-04
1/1 [==============================] - 1s 1s/step

GRU Model Evaluation:
Model Evaluation Metrics:
RMSE: 4050.370
MAE: 2701.552
MAPE: 4.64%
R-squared: -0.598
Explained Variance: -0.574

LSTM Model Evaluation:
Model Evaluation Metrics:
RMSE: 4064.404
MAE: 3084.398
MAPE: 5.21%
R-squared: -0.609
Explained Variance: -0.585

Of course some Explanation

  • Epoch X/300: Current training epoch out of 300 total epochs. One epoch is one complete pass through the training data.
  • Loss: Training loss value at the current epoch, indicating how well the model fits the training data.
  • Val_loss: Validation loss, showing how well the model generalizes to unseen data.
  • LR: Learning rate, a hyperparameter that controls the step size during optimization.
  • ReduceLROnPlateau: A mechanism that reduces the learning rate when the validation loss stops improving.
  • RMSE: Root Mean Squared Error, a measure of prediction accuracy (lower is better).
  • MAE: Mean Absolute Error, the average absolute difference between predicted and actual values.
  • MAPE: Mean Absolute Percentage Error, a percentage-based error metric.
  • R-squared: Indicates the proportion of variance explained by the model. Negative values suggest poor model performance.
  • Explained Variance: Reflects how much of the variance in the target variable is captured by the model (closer to 1 is better).

That's a wrap, folks! 🎬✨

Remember, a whole bunch of super-smart people 🧠 way brainier than lil' ol' Notoko put a ton of effort into those predictions! 💪

So give yourselves a pat on the back, 'cause you're part of something amazing!

Disclaimer

*The content provided in the Crypto Price prediction section is intended for informational and educational purposes only. It should not be construed as financial, investment, or trading advice. The opinions, analysis, and forecasts expressed in this section are those of the authors and do not necessarily reflect the views of the entire Bitconometrics team or any associated entities.
The information provided on this cryptocurrency blog is for educational and informational purposes only and should not be construed as financial, investment, or trading advice. The authors, contributors, and administrators of this blog are not licensed financial professionals and do not hold any formal qualifications in the fields of finance, economics, or cryptocurrencies

Cryptocurrency markets, including Bitcoin, are highly volatile and can fluctuate significantly in a short period. The information presented in this section is based on historical data, current market conditions, and the authors' interpretation of various economic indicators. Although we make every effort to provide accurate and up-to-date information, there is no guarantee that any analysis, prediction, or projection will materialize or prove accurate.

Before making any financial or investment decision, we strongly recommend that you consult with a qualified financial advisor, conduct your own research, and assess your personal risk tolerance. Bitconometrics, its authors, and affiliates will not be held liable for any direct, indirect, or consequential loss or damage incurred as a result of the use or reliance on the information presented in this section.

By accessing and using the content in the price prediction section, you agree to assume full responsibility for any investment or trading decisions you make based on the information provided herein.*