how to predict what will happen with the price using machine learning

[bitcoinalexis]

Today I’m sharing an app with you that has already made more than 35% so far this November of the current year, as of today Nov 24, 2025.

It analyzes USDT pairs listed on the ByBit exchange. It’s completely free and does not require registration or any personal data (exchange API, etc.).

It uses machine learning and statistics, applying a mathematical formula on the order book along with technical analysis. If you’re wondering how I’m able to do all of this: I studied a bachelor’s degree in economics, and I am currently studying another economics degree focused on data analysis and machine learning. I apply what I’ve learned at the university plus my programming knowledge, which I’ve had since I was 13 years old up to now.

Today I released an update that helps us know when a pump has ended.

First, the app will show you the signal of an asset that is starting a bullish trend or has experienced a pump.

How to earn returns?
For example, if it detects a pump in “ORDERUSDT”, I go to TradingView and check if the MACD value on the 1h or 4h timeframe is already negative. If it is already negative, the price tends to drop by about 3%–8%.

You can find an example by clicking the help button at the top of the pump signals window within the app. Any questions, comments, or suggestions are welcome, guys. I hope you like my project. Regarding ads: the app is just an interface — all these tasks are executed by Python scripts I have hosted on Railway, and the ads are a way to support me so I can continue paying for the servers.

Best regards, and may you earn well.  

Link app:
https://play.google.com/store/apps/details?id=com.angylabs.angysignals


I want to let you know that I have documented my results and the work I did on my YouTube channel, which I’m sharing with you. I’m a Spanish speaker, and the content is in Spanish.
https://www.youtube.com/@BitcoinAlexis/streams

I’m sharing it in case anyone here understands Spanish.

[Aanuoluwatofunmi]

As long as am concerned, this 35% result achievement for the month of November and considering only few days are left shows that it's almost impossible to achieve up to 50% and also, this shows that there will be times we can have more less, which makes it more advisable that we should learn to do everything by our own self, because there's no accuracy in what we can get from engaging any of these developed apps for better results in trading, than doing it by our own self.

[BitMaxz]

What strategy does it use? I tried the app, and it is giving me a signal to go sell with an entry, target, and SL. The risk ratio is high, and it is targeting a small profit.
So, I'd like to ask what strategy or how exactly it works. Is this based on moving average?

I check the pump detector; it looks similar to the top gainer on the Coingecko and Coinmarketcap apps, and it crashes if you try to tap anything on the pump detector.
Plus the interval of ads is pretty short; it keeps showing after a few uses.
Honestly, I'm against any signal app; I am more confident in analyzing the market on my own.

[bitcoinalexis]

What strategy does it use? I tried the app, and it is giving me a signal to go sell with an entry, target, and SL. The risk ratio is high, and it is targeting a small profit.
So, I'd like to ask what strategy or how exactly it works. Is this based on moving average?

I check the pump detector; it looks similar to the top gainer on the Coingecko and Coinmarketcap apps, and it crashes if you try to tap anything on the pump detector.
Plus the interval of ads is pretty short; it keeps showing after a few uses.
Honestly, I'm against any signal app; I am more confident in analyzing the market on my own.

https://github.com/bitcoinalexis/AngyApp/blob/main/HOW%20TO%20USE%20MACD%20WITH%20PUMP%20DETECTOR.png?raw=true

Hello my friend, thanks for your comment and contribution. By the time you read this, I’ve already fixed the bug in the Pumps section — the app no longer closes. About “Ads,” thanks for your feedback; I took it into account and added a limit to the number of ads displayed, so they’re no longer annoying. I invite you to update the app.

The version that includes these changes is 0.1.2. If it’s not available yet, you’ll need to wait for the Play Store to approve the update.

“Pair Analyzer” — this tool is meant to support you when you are doing technical analysis of the pair.

How does it analyze?
It takes into account, in real time, executed trades + the order book (I won’t explain the Coca-Cola formula either).

In the “How to use” section, usually when you run the analysis and the confidence value is >83%, it’s very likely that the price prediction will be accurate. IMPORTANT: you must keep an eye on the pair. It has happened to me that the trend later changes — a SHORT signal can switch to LONG and vice versa.

Now, about the pump detector — for example, a Pump alert:
“MELANIAUSDT”

I use the MACD indicator. When the MACD indicator is already in a negative value and closes in a negative value, it tells me that the bullish trend is over, and the “Pump” is also over.
https://github.com/bitcoinalexis/AngyApp/blob/main/Indicador%20MACD

Screenshot explaining the example:
https://github.com/bitcoinalexis/AngyApp/blob/main/HOW%20TO%20USE%20MACD%20WITH%20PUMP%20DETECTOR.png

Regarding personal trading, I have two goals: to reach the audience that likes automated trading, and to reach conservative traders like you and me, offering tools that support us when performing technical analysis. Many tools are coming in future updates.

If you have any comments, I’m keeping an eye on them.

[bitcoinalexis]

As long as am concerned, this 35% result achievement for the month of November and considering only few days are left shows that it's almost impossible to achieve up to 50% and also, this shows that there will be times we can have more less, which makes it more advisable that we should learn to do everything by our own self, because there's no accuracy in what we can get from engaging any of these developed apps for better results in trading, than doing it by our own self.

Hello my friend, thanks for commenting. The performance percentage is variable.
I was busy with university and couldn’t do much trading. If I had had more time, I have no doubt the percentage would be higher. The performance each user generates is variable since this tool is manual-use. In the future it will be automated, just like the existing tools I’ve already created and shared on my YouTube channel.
Steps for the development of my tools:
1 - Create an alert system
2 - Review alerts and look for patterns in trends, and filter how to operate with the obtained data
3 - Create the tool
4 - Use the tool manually for 30 or 60 days in volatile markets
5 - If the results are good, create a script so the tool becomes automatic (trading bot)
Greetings my friend, thank you very much for your comment, your contribution, and for testing the app. Your feedback helps me improve.   

[adultcrypto]

I don't know how to verify this because I have no expect knowledge of machine learning, but what I have to say is that this will not be different from the numerous trading bots we have out their. You claim you have made 35% this November, if this is consistent and sustainable, I'm confident you will hit major clients who will be willing to invest millions into your software after doing their due diligence. I will be monitoring this thread to see the progress you are making. Assuming you are reputable member of this forum, it would have been easier to download and try your app.

[Cryptohygenic]

If you like, batch the highest degree in the world I don't believe in any inventory that claims to be reliable trading bot.
Furthering your economy education as a professional economists is indeed a good one, maybe you can be employed in the economy sector so that you can contribute to revive the global crashing economy, but I bet you will have no way to handedly manipulate the volatility markets with whatever tool. There are also professors in the universities who has tried to analyse the market but fails and give up after realizing the market is unpredictable and no significant of strategies can be reliable when determining it future performances.