Author

Topic: ✅🔥⚡🚀[ANN][BETA] ⭐ COINFARCAST⭐ [ETH, Ripple, Stellar] Scientific Forecasting (Read 37 times)

full member
Activity: 299
Merit: 100
Ethereum ETH 7 days forecast

An Econometrics model built using 273 days historical close prices.


Back testing results

Average Price Deviation: 1.5%
Correlation Coefficient R2: 0.65
Direction Accuracy: 0.83

full member
Activity: 299
Merit: 100
Please let us know. what other coins would you like us to add?
full member
Activity: 299
Merit: 100
Coinfarcast, your market analyst.


Simulate the market, back test your model and predict prices in 3 mouse clicks




Coinfarcast.com provides state-of-the-art analytics using Big Data and AI with a focus on market sentiment analysis.

Why are we different

Coinfarcast allows you to combine extensive data about the market in one mathematical formula

Market Psychology Big Data

Available
•   Millions of traders’ reactions on social media.

Planned
•   Global and market news feed and articles.
•   Experts’ price analysis articles.


Extensive market data

Available
•   Tens of technical and fundamental indicators.
•   Minute by minute historical price and volume history.

Planned
•   Stock market indicators and commodity prices.


Scientific, objective & straightforward
We fight subjective & pseudo-scientific analysis

•   Transparent prediction back-testing.
•   Advanced prediction accuracy statistics.


How it works

1.   Choose your model settings.
2.   Calculate a mathematical model.
3.   Evaluate model back-testing results.
4.   Get model prediction accuracy.
5.   Use the forecasts or create a new model.


How to read modelling results

Don’t take predictions from any model. As the saying goes “All models are wrong”. You are only looking for an “Adequate” model.

1. In the “Price deviation” graph, don’t consider any model until you find small and consistent deviation. The graph calculates how far the predicted prices deviate from the actual prices. Below 2% deviation is small.


2. Directional accuracy is a value between 0 and 1. Best is 1 while above 0.66 is just acceptable.

3. Correlation should only be used if the Actual (not the predicted) price movement is more similar to a straight line (So, linear). The higher correlation, the better.

4. Average price deviation: this is an average of the “Price Deviation” chart.
Take the forecasts or create a new model.

5. Only when you have an “Adequate” model, you can use the forecasts of this model while expecting a deviation equal to plus/minus average back-testing price deviation.



We Need Your Opinion

We are currently in a Beta test from Apeil to June 2021. Some features are free and we will open the other features for later for a testing period. We ae looking forward to hear your opinions about the service.

Join us on Telegram


Linkedin page





Jump to: