Can Machine Learning truly predict Bitcoin's price?

[Yamifoud]

Or maybe, we need to buy and use one of these AI/machine learning in order to evaluate its performance. This is gonna be the best solution and to prove something, I think.

It's useless and it could be wasting your time and money, there's no such way to predict the future price of Bitcoin.
But one thing all we know is that Bitcoin price was always ended at a high price.  Even though this machine has an AI algorithm system that can able to predict but I think it isn't accurate.

What kind of machine it is?
If so, there could be a chance it will use in an automated trading bot, and this may be the perfect tool in trading.  But yes, for sure the performance doesn't have a guarantee that it has an accurate result.

Indeed, a waste of money and time. And there is no way to predict the future of Bitcoin and I think, we've done with our own market analysis and TA which I think we don't need such machine.
I don't know how the invention is made and how it proves that it predicts the right price. We keep thinking about the volatility of the market which makes these predictions and any form of predictions will never be perfect or even too close to it.

[xSkylarx]

This is like an EA in trading, where they are using a bot, but the only problem is that it is a program. My opinion with this is that this AI was also programmed to check the history. This is mostly the strategy with this and also the probability that it would happen again. Also, the candle patterns and technical analysis that a trader would use and calculate the probability that it would happen, but still it is not 100% that would predict the price.

[Reatim]

This is like an EA in trading, where they are using a bot, but the only problem is that it is a program. My opinion with this is that this AI was also programmed to check the history. This is mostly the strategy with this and also the probability that it would happen again. Also, the candle patterns and technical analysis that a trader would use and calculate the probability that it would happen, but still it is not 100% that would predict the price.

There will never be a 100% in prediction mate, even how good and accurate yet since it is Prediction this will never come perfectly , instead i believe that this will only give hints and opportunity to be in specific coin but not to come in permanent situation.
so Yes it can predict the price roughly and not completely .,

well, how can they predict that exactly, whereas bitcoin or crypto trading is a trade about demand and need. well, actually, looking at it from youtube, I can't believe it 100%, moreover it's a machine. even bots that are designed to have a fairly accurate prediction system can't even predict bitcoin price movements. maybe it can give you speculation about from the information obtained, but I'm not sure that it's 100% true.

and also has some sort of manipulation inside.

[Akash1243]

As you have mentioned the crypto market cap is way smaller compared to the stock exchanges of even developing nations and hence has a whole lot of whale manipulation,even sound technical and fundamental analysis are thrown out of the window when the whales enter the picture,hence any ML AI algorithm would have a much greater chance at predicting stock indices over crypto in general.
Having said this i still believe that even if in the future we get an insane over the top algorithm developed if it gets popularised too much or a large capital is invested based on the algos opinions then it would create a paradoxical situation,i.e. lets say you are a wealthy investor and your algorithms prediction is a price increase ,if you invest a huge amount now you are the one who is actually being a catalyst for this growth.
But right now there is no algorithm which even has decent accuracy in predicting prices of stable assets let alone cryptocurrencies.

[franky1]

As you have mentioned the crypto market cap is way smaller compared to the stock exchanges of even developing nations

that can be fixed. in just a couple days.
make a altcoin with 50trillion coins. sell just 0.001 of a coin for $1.
making the altcoins market cap 50quadrillion dollars

all for the price of $1 sell

what you then learn is that market caps are a meaningless statistic to look at
what you learn is stock market cap is a meaningless number

stock prices are not based on some large pool of dollars held to match every share/stock.. its based on the current price of an order (of any % of a stock) and this price is then with no real dollar involvement multiplied by how many stock/shares there are

crypto market caps have no dollar reserve either. its just based on the current price of the last order(even if its decimal) and multiply that price by the coins in circulation.

so you can change the market cap of bitcoin by large amounts of $measure. without having to see any large amount of dollar spent

[barbara44]

Personally, I don't think any ML or AI tech can make any price prediction now because we have a lot of whale manipulation (because the market size is smaller than any other market). Well, while I may be wrong I would like to know what you think. Have you ever gotten a satoshi richer by following the prediction of an algorithm?

In my understanding, science has evolved only up to the level of machine learning and artificial intelligence is still not existing for all of us; coming back to the capability about machine learning to predict the bitcoin prices; I guess machine learning is not having big differences from what we do use for trading in the name of technical analysis. If machine learning is programmed in a versatile way then it may become capable of predicting the market up to some extend.

So, we cannot completely ignore the potential of machine learning process about predicting the future of bitcoin markets as machine learning must be an enhanced version of technical analysis in somehow automated way.

[Vaculin]

No they can't predict the price of Bitcoin is unpredictable no one could predict what will be the price of Bitcoin in the next day or a week or even a month.

Maybe LM and AI can create small chances in its first uses but in the long run, it will only create false predictions and whoever believes on it might only see himself losing in the end. Because i believe that the crypto market is naturally unpredictable and that bitcoin is very volatile so there is no way that any certain thing will be able to predict the price correctly. And seeing that bitcoin and altcoins' prices can still be manipulated by big whales, and any sudden positive or negative events will affect the whole crypto market. For now, i believe its more on the market sentiments that create the price of bitcoin and altcoins.

[LangePara]

Goodluck with ML, one splash of a whale and its gone

[AmoreJaz]

No they can't predict the price of Bitcoin is unpredictable no one could predict what will be the price of Bitcoin in the next day or a week or even a month.

Maybe LM and AI can create small chances in its first uses but in the long run, it will only create false predictions and whoever believes on it might only see himself losing in the end. Because i believe that the crypto market is naturally unpredictable and that bitcoin is very volatile so there is no way that any certain thing will be able to predict the price correctly. And seeing that bitcoin and altcoins' prices can still be manipulated by big whales, and any sudden positive or negative events will affect the whole crypto market. For now, i believe its more on the market sentiments that create the price of bitcoin and altcoins.

the fact that there are so many factors in play in btc market, even if we say AI already integrated those factors, still, there will be other factors that may possibly influence the market. it would be hard for them to consider all those factors and know the contribution in the market. if in case, someone will come up with such algo (at least near to reality), do you think he will share it to the public? i don't think so. he will utilise it as long as he in profit and just disclose when his algo is not working anymore. but up until now, i don't think someone already devised such algo

[Questat]

I was thinking about how it works as I'm not convinced of its effectiveness knowing how volatile is the market. Yes, of course, it was going to believe that it could give some predictions as it was used to be but the problem is that if that was close to it. If that going to happen, then people would love to try it. For now, as it wasn't proven yet that it really works great, that is something in question and possibly people will just forget such an invention but rather use their own TA.

[NelfiNovita]

I've watched a lot of YouTube videos and read a lot of articles that claim that you can predict the Bitcoin price with Machine Learning and artificial intelligence like using some algorithm. Truth be told it is both intriguing and almost at the same time confusing. Personally, I don't think any ML or AI tech can make any price prediction now because we have a lot of whale manipulation (because the market size is smaller than any other market). Well, while I may be wrong I would like to know what you think. Have you ever gotten a satoshi richer by following the prediction of an algorithm?

Someone's predictions are not always correct, let alone using tutorials such as youtube, algorithms and sophisticated technological systems. Now the price in the market can fluctuate so it is difficult for us to predict the market price will go up or will go down, besides that the big whale also plays a role with the current market price.
I myself can't predict market prices using an algorithmic system and so far I can only use the theory of holding in the long term so that I can get big profits.

[ranochigo]

Some misleading responses here.

I've worked with machine learning for stocks prices, which by all means are far more consistent with the general trends than Bitcoin prices. I used Colab with Adam as an optimizer for the weighted moving average of the stock. Machine learning CANNOT accurately predict the prices of the stocks, solely because there are a magnitude of factors that could affect the market movement. Specifically, in my case, Covid-19 skewed the data significantly and resulted in up to 40% lower predicted prices vs actual. It would work if the stocks follows the typical market cycle, which it doesn't.

Now, to Bitcoin, it is even worse. The problem is that Bitcoin doesn't follow any trends, it is affected by various factors which are often unpredictable (govt regulations, stocks market). If you were to plug in the data for 2010 to 2019, you would have a huge deviation as compared to the actual prices.

[ChiBitCTy]

First off AI that exists today isn’t really that amazing compared to what we believe and almost assuredly know what Artificial Intelligence will be one day. It’s not self learning which is what future AI will be mind bogglingly able to do (and scary for that matter). But it’s forever going to be impossible to predict because you simply can’t read peoples minds, you don’t know when they plan to buy or sell (outside of stuff like futures and options).

[franky1]

First off AI that exists today isn’t really that amazing compared to what we believe and almost assuredly know what Artificial Intelligence will be one day. It’s not self learning which is what future AI will be mind bogglingly able to do (and scary for that matter). But it’s forever going to be impossible to predict because you simply can’t read peoples minds, you don’t know when they plan to buy or sell (outside of stuff like futures and options).

actually.. with enough data it does become easier,
for instance if you have access to an exchanges user database+their balances+their order history. you can see which users just buy the dips. which users FOMO(buy the hype). which users DCA(just buy at X day of month no matter the price)
you can the group people of the 3 trade types. total their combined holdings per group. and see how much combined yield/power over the market they hold in each category to see how much the market can move by when each category scenario triggers.

EG if you know, for instance that the 25th of the month is a high total yield day for the DCA group. then thats more buyers buying on the 25th than other days so expect more buy pressure on the 25th

patterns can be found... IF you have the data.

[B-Bit]

The robot predicts the price of the coin through a period of analysis to predict whether the price will rise or fall. But I can't fully believe the robot's predictions. The best way is to learn to analyze the market by yourself, and it is better to analyze the price of coins based on the heat of the market. The robot cannot predict what the giant whale will move. So it's better to rely on yourself.

[wxa7115]

Machine learning is not exactly easily to implement - you need high quality data and you need to configure the network in just the right way so that it would generate the best model, but it would be in this situation mostly just guessing on how to do this. And of course neural network can't get 100% accuracy, it probably can't even do 60% accuracy, but anything over 50% is a success, because it means that profit can be generated over a long period of time.

But this doesn't mean that you should go buy one of those Bitcoin trading bots, most of them are likely just scams, because if someone had a successful bot, they would just use it themselves, instead of sharing it with others and reducing its performance.

And another weakness of machine learning is that it can't react to real life events. Like how Bitcoin crashes or pumps when certain big news happen (China ban, Tesla buy, etc.).

It is true that it is not something easy to do as the data needs to be accurate in order to produce a model that can predict what the market will do in a day to day basis, but it is also not something as impossible as it may seem at first sight.

So at the end whether a person can implement something like this will depend entirely on their abilities when it comes to this particular branch of artificial intelligence and whether or not their expectations of the results they can get are realistic enough.

[kawetsriyanto]

The machine can make a Bitcoin price prediction, but not sure how good the accuracy of its prediction is. I mean even an expert is difficult to predict with good accuracy, how can a machine do a better prediction? We must remember that making a price prediction isn't based on the price chart history only, we must learn how the current news about Bitcoin and crypto market. A machine must have some limitations. It may be programmed to predict the next price based on the chart history only, while Bitcoin price doesn't always repeat the same scenario in the past. Sometimes price change is unpredictable and depends on the current news/issues in crypto market. So, I doubt if a machine can predict Bitcoin price properly.

[DoublerHunter]

The robot predicts the price of the coin through a period of analysis to predict whether the price will rise or fall. But I can't fully believe the robot's predictions. The best way is to learn to analyze the market by yourself, and it is better to analyze the price of coins based on the heat of the market. The robot cannot predict what the giant whale will move. So it's better to rely on yourself.

^ Definitely right, there is no need for a machine to predict the market price.
With a combination of technical and fundamental analysis you have a preference to predict the market, probably don't have an exact result but I think that is not far. I don't believe in the machine algorithm because they usually have mistakes and the fundamentals sometimes where it is silently will strike without any tools to predict. Doing your own research will probably be a good idea and it is totally free.

[BuNga_cute]

The machine can make a Bitcoin price prediction, but not sure how good the accuracy of its prediction is. I mean even an expert is difficult to predict with good accuracy, how can a machine do a better prediction? We must remember that making a price prediction isn't based on the price chart history only, we must learn how the current news about Bitcoin and crypto market. A machine must have some limitations. It may be programmed to predict the next price based on the chart history only, while Bitcoin price doesn't always repeat the same scenario in the past. Sometimes price change is unpredictable and depends on the current news/issues in crypto market. So, I doubt if a machine can predict Bitcoin price properly.

Now many sophisticated machines have sprung up, because technology has developed very rapidly. So I believe machines can predict the price of
Bitcoin, but the problem is that the prediction results can't be 100% accurate. Because it is impossible with volatile Bitcoin price movements there is
a machine that can accurately predict Bitcoin prices. If there was such a machine maybe its creator would not have sold it to anyone else and would
have used it himself. But it is impossible for a machine to predict 100% accurately, because as you said it is not only studying the history of Bitcoin
price movements to be able to predict Bitcoin prices. There are other factors that influence it, moreover all human creations must have limitations.
That's why I never 100% trust trading bots, because it's not easy to predict Bitcoin prices.

[GeorgeJohn]

Let us look at this from our own  understanding of cryptocurrency generally, we actually known from the origin of cryptocurrency and more especially bitcoin that no machine at this era that can predict the actual price of bitcoin perfectly and i don't believe its in existence, the price of bitcoin is been determined from the market structure of demands and supply of cryptocurrency, the commodity is what we consider first from cryptosphere, and i believe vividly that machine can't predict the price of bitcoin but it only trade a coin.