Topic: Can Machine Learning truly predict Bitcoin's price? - page 4. (Read 549 times)

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.

These machines are programmed to make predictions, then it's possible they could make it but the question is if that is realistic? I don't think how reliable it was, we know these machines are not too good compared to the human brain and we know that nobody was able to predict the right/exact price of Bitcoin, so how much more for these machines?
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.

I think there are some things we need to take into account, when someone does technical analysis that person can have a chance to predict the price but it only works if it's a short amount of time like a 4h or daily or week chart; but it is very difficult to predict with precision, if there are people who do technical analysis then it is not possible to have Machine Learning that does this service better, Has anyone seen any Machine Learning in action doing price forecast?

But watching what you don't know from video and make a frequent practice on them can equally make someone perfect at some instances, because i a friend of mine doesn't know what is trading in the year 2015, but from the year 2017 he started using internet to inquire for online trading tutorial, only one year practice of this trading through YouTube videos it's now a master in trading industrial, i think i refute such concept that watching a YouTube video will not make someone perfect. Even manual lectures too can't make someone perfect, the only thing that's involve is concentration.

I don't believe the price of bitcoin can be known for sure by only using machine learning and AI because changes in bitcoin prices are purely based on supply and demand made by humans in the market. machine learning and AI can only predict bitcoin prices based on market habits and the accuracy is not that great.

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.).

They can.

But the question is if they're going to be that accurate with that prediction? IIRC, I've read something like this before about predicting the future price of bitcoin. The same thing, they did predict but they're not just as accurate as they are.

There are unprecedented things that would come to the market and they can't predict that. So, if it's just about predicting the price, that's easy.

And to tell you, not every youtuber is a good content creator to watch.

Understand that, not every video you see on Youtube that you have to believe the information Youtubers share because Youtubers make hype based videos for their own greedy purpose.
Having said that, there's a chance that machine learning will predict the price of  Bitcoin but that doesn't mean the price it predicted is 100% sure because we are dealing with a volatile market and no one or thing can predict an accurate result of the market price.

Some machines work, while some do not.

Eventually in the long run, as you feed more information on the machine, it will only give you the answer based on that information, and I don't it would precisely predict the future movement because if it will, then probably trading will be so easy and making money is not a problem. I suggest using it only as one of the references, still, look for more current news and information as they likely are a more reliable basis for predicting the market movement.

What kind of machine?
There were actually applications or program created to detect buying and selling of crypto which is illustrated through a chart to predict the next Bitcoin price, but that's the only thing that can make the closest possible price prediction.
There are no other thing that can actually predict Bitcoin price without detecting the buy and sell volume of such asset.
 
Unless that machine you're talking about is a time machine.
Well, that can accurately predict Bitcoin price lol.

Yeah, of course, however, you can expect that it will give consistent results as crypto by nature is very unpredictable. Machines do base their prediction on past movements and trends, but how about the news which is a major reason why the market moves? I mean the latest rumors, and the FUD, a machine could not detect that.

The first aspect to determine is what is the event to predict. Obviously it’s price here, put that is tied to a given timeframe. The algorithm therefore needs to aim at predicting BTC price in that given timeframe (i.e. price in the next 5 minutes, next hour, next week, next day, and so forth). Each of these timeframes constitutes a different objective, and it represents the starting point to derive a set of algorithms that try to predict the value of BTC in that timeframe.

Price depends on many factors. The simplest set of algorithms will work with historical price data; the more granular the better. If there is available information on trading volumes, and order books, all the better. That should be suffice to do something similar to TA. Nevertheless, there are many more elements that conceptually should be thrown-in to the bundle in order to be processed as data inputs. The basis is having that data tabulated in a precise and timely manner in a historical set of data.

We often perceive that market is very much influence by fundamentals; or rather, fundamentals act as catalysts that are then exploited by those who can in the direction that they will (or not …). On top of that there are plenty of minor news that play their role, psychology of numbers, sentiment, fomo, fud, derived products that may trigger a BTC price sway, or are triggered themselves in price and then they trigger BTC price, BTC tenant composition and weight (i.e. whales, institutions, us), and so forth. These factors are very difficult to factor in to the models in their accurate and precise measure. There’s probably a fair share of NLP related in the process of attempting to ponder them into the prediction.

Can all the above be taken into account into a model? To a certain degree for sure, but precision in the output (again, for a given timeframe) requires the inputs to be as precise as possible, and their influence identified in scope and time. I’m sure that there are many attempts at the above, but rather sceptical on the precision of outcome (better chances on the tendency).

I recall reading this some time ago (3 years now):
https://towardsdatascience.com/using-recurrent-neural-networks-to-predict-bitcoin-btc-prices-c4ff70f9f3e4

The article concludes the following:
And that was just working on historical data, without attempting to work on all the rest mentioned above.

I don't have much of technical knowledge about the algorithm and related stuff. Based on a layman's view I find it to be impossible. If that is possible by now corporate firms and other richest could've got into development of the machine learning through which wealth can be multiplied with ease.

If such a machine learning exist, then the market won't be true as now and this will make decentralised system to be centralized which is against the objective of Satoshi.

Like most of the time it depends on what are your expectations out of the predictions that can be produced by Machine Learning, if you are expecting that the markets can be predicted all the time with perfect accuracy then it is obvious nothing can produce results like that.

However can the markets be traded that way and produce profits? Of course, and there are many trading firms doing exactly that, however if you do not know anything about coding, computers, artificial intelligence and machine learning then you better stay away from it.

There are many scopes and aspects that need to be considered if you want to develop an algorithm to predict bitcoin price. I didn't know which video or article you refer to but if they claim they can predict the price without showing proof then it is doubtful. If they truly can predict it they will disclose the algorithm and benefit it from their own, anyway.

After all, the price isn't determined solely based on past history or any correlation metrics that are used to manually predict the bitcoin price, it is vague. And to determine which data is needed to make the algorithm/ML works is complex. So, as of now, I think ML can not predict the price.

The price of bitcoin is highly dependent on many unpredictable variables, offline/online events and happenings so much that the outcome will be impossible to predict accurately, speculation can only be made even by AI's. IMHO, AI's can only give possible predictions based on current events but not future events that are unplanned.

How would the machine able to analyze each and every news that there is on the internet and how would they be able to track the whales and their wallets as well. I don't that that would work. They would also need a particular market size which is always changing. The volume of Bitcoin, the supply, the volatility is all against any software learning the patterns. There are softwares out there in the market which tends to make the users believe that they can predict the market price therefore at the end of the day I think :
A software will only be able to predict something which is backed up, based on some equations, is predictable not something like bitcoins which is 50% controlled by whales but 50% controlled by individuals, which is a very unpredictable equation.

Any other market? Man. 1 daily bitcoin volume = 150 daily volumes on whole stock market in my country across all 500 available stocks.
There are only few markets bigger than bitcoin's in terms of volume and liquidity but none of them compete with bitcoin if you add volatility into calculation.

Can AI predict bitcoin price? For sure it can in middle/short term. Manipulation you are talking about happens from time to time and is nothing else than statistic error because AI bot can trade 24/7.

As someone whos working as a python programmer also been studying Ai and ML during the last years I saw many people claim they created a bot that will predict the price of bitcoin, gold, etc... and it will give amazing profits in long term I tried to create one and used many systems such s linear regressions but the but didn't work out for me. All I can say is at the current technology level in the world, the bots are unable to do such a thing and using but for this purpose can be very much risky, however, AI is the future and soon we will be able to use AI and ML to predict the price of anything and this is just one small part the AI world is the future.

In most cases, traders face some common problems during trading. Whatever, Long term trader or short-time trader both become emotional during trading. In fact, the number of emotional traders is higher among us. For doing manually analysis with open eyes you need to gain a lot of knowledge and experience about trading and the stock market. Without sufficient experience or proper knowledge, you will not be able to reach any good conclusion by analysis. Trading bot can easily do the analysis within a short time for the traders, in this case, it is very efficient. And since bots predominate over technical analysis and provide emotionless predictions, the risk of loss is relatively low here, I said "relatively". Giving prediction is not its main work, it is subsidiary work And trading bots are helpful for traders' to do technical analysis.