Last Updated on February 18, 2021 by admin
These predictions are available for all of our membership tiers, including the free plan.
This post will show you how to get the most from them.
As a crypto trader and investor, you can use our crypto price prediction indicators to supplement your trading strategies, including the Trade Signals from Crypto-ML.
Here is a quick overview of how to interpret the results:
|Positive Prediction||Negative Prediction|
|Normal Market||Bullish: Follow Predictions||Bearish: Follow Predictions|
|Bullish Manipulation||Strongly Bullish||Bullish|
|Bearish Manipulation||Bearish||Strongly Bearish|
The Member Dashboard displays 6-hour and 12-hour price predictions:
In this example, you can see that in 6 hours:
- BTC is predicted to go up 0.52% within the next 6 hours. In 12 hours, it is predicted to go up only slightly more.
- ETH is expected is excepted to move up slightly by 0.32% in 6 hours, but is predicted to go up 1.19% in 12 hours.
In addition to price prediction, Crypto-ML machine learning also identifies when the market is not following traditional patterns. This occurs rarely, but can result in huge market movement.
Machine learning predictions, and statistics in general, require sophisticated methods of interpretation to get the most from them. Our case is no different.
The “Manipulation Detector” section shows you how to interpret the predictions.
Manipulation Detection States
If the Manipulation Detector is gray, then the predictions are likely to be accurate.
If the Manipulation Detector is green, then exceptional conditions are happening that are pushing the market higher than expected for current patterns.
Price will likely go up more than predicted.
If the Manipulation Detector is red, then exceptional conditions are happening that are pushing the market lower.
Price will likely drop more than predicted.
Manipulation Detection Details
Statistically speaking, our predictions are very accurate. Most of the time, the manipulation indicator will be gray.
Lower on the Member Dashboard, you will see a couple of key metrics for our models: R-Squared and Mean Absolute Error.
As a simplified explanation, this roughly means that the model for BTC-USD in this screenshot is accurate within a +/- 0.33% range 94% of the time. This metric is determined by testing predictions on data that was not used for training.
Said another way:
If our prediction says +1.5% in 6 hours, then 94% of the time, price will actually move to between +1.17% and 1.83% in 6 hours.
However, this also means that 6% of the time, our predictions are wrong. Normal market patterns are not describing market behavior. Some non-standard force is manipulating price.
The fact that normal patterns are breaking is interesting and acts as a key trading signal. When this happens, you stop trusting the predictions and instead begin trading in the direction of the market manipulation.
Learn more about our Manipulation Detection.
Learn How Crypto-ML Works
The prediction values you see on the Member Dashboard are a direct output of this step in our machine learning flow:
If you’re interested in learning more about how these predictions are generated, you can dive into additional detail about our machine learning workflows. Learn how Crypto-ML works.
Note: There are many different ways to approach this, which you can review in our do-it-yourself Tutorial: Bitcoin Price Prediction with DIY Machine Learning in Excel.
Crypto-ML provides machine learning for crypto traders and investors. Gain crystal-clear signals and deep market insights. Add predictive capabilities to your toolbox. Learn more and join for free.