User Guide — Alphacat Cryptocurrency Real-time Forecasting Series Tool

Based on a prediction algorithm built using the TensorFlow artificial intelligence framework, the Alphacat digital currency real-time forecasting series uses the PRNN-LSTM* algorithm to predict the trend of a digital currency in the future timeframes of 15, 30, and 45 minutes, while also drawing three possible trend lines.

The middle curve shows the estimated average price in the coming 15, 30, and 45 minute time-frames, while the upper and lower curves depict a 60% possible price change range, that is: the probability that the future trading prices will fall in this range will be 60%. The two circles at the bottom of the page will present to you the complete forecast. The user can clearly see the AI engine’s prediction for the digital currency price trend.

Here are some use-cases for your reference:

1. Helps users decide when to buy and sell

Through research and analysis, when a user wants to buy a digital currency, should they buy it now? Users can use the Alphacat predictive tool to support their buying decisions. If the user see that the predicted trend line is obviously rising upward, and the upper and lower trend range lines are narrower, the user has a strong indicator that buying is an option. On the other hand, if the trend line is falling downward, the user can wait for the trend-line to change into an upward motion, or they can use market tools to short their position to make gains on the down-trend.

Using the same logic, users who want to sell digital currency can wait until the trend line is predicted to go down.

2. Helps users compare the risk factors of different digital currencies

Users can compare the results of the multiple predictions of two different digital currencies. If the average width between the upper and lower predicted range lines of digital currency A, (for example, the width is about 21% of the price of the currency), are significantly larger than the average width between the upper and lower predicted lines of digital currency B. It can be concluded that the risk of investing in digital currency A is much greater than that of digital currency B.

3. Use the results of real-time prediction as trading signals

Users with rich experience in digital currency trading and a high level of trading skills can use the results of real-time forecasting as trading signals to systematically trade and obtain long-term profit. First, the user needs to choose the transaction currency, then decide how much money to trade. The user then would make an independent judgment based on the general environment of the market. They would then trade according to that currency price trend, [if it’s relatively stable or rising, (for example, the price of the currency is above the MA20)].

Users can continuously track predicted trends. If the forecast trend line is obviously rising upward, and the trend range is narrow, and the rising probability, (displayed below the predicted graph,) is higher than a certain threshold (such as 65%), and this position holds for a certain period of time, (such as ​​after 3 hours). Or when the “Rise Probability” shows below a certain threshold, (such as 45%), the user can decide to sell.

Since the currency price trend fluctuations are very random, the forecasting tool provides a probabilistic forecast; sometimes the forecast is accurate, sometimes it can be incorrect, and the result of a transaction could be profitable or could end in a loss. Only after making multiple transactions will the overall win or loss reflect the advantages of Alphacat forecasting and make user transaction records profitable.

We hope that upon reading this article you will have a deeper understanding of the characteristics of Alphacat’s real-time forecasting tools, and can better use the tools to help you make trade decisions. In the coming series of articles, we will continue to introduce the functions of the real-time forecasting series tool.

*(PRNN: Pipelined Recurrent Neural Network; LSTM: Long Short-Term Memory; the LTSM neural network model, suitable for time-series prediction, was combined with the PRNN algorithm to obtain the PRNN-LSTM algorithm).

This information is neither a recommendation, nor an offer to sell or buy, nor a solicitation of any type to purchase or to conduct any investment.

You should assume full responsibility of your own profits and losses.

The Alphacat team will make the final interpretation of these rules.