Drop a like for more machine learning versions of the top indicators!
I know that I totally didn't keep the video short like I said, but luckily the maximum time for a video is 10 minutes so you didn't have to listen to my blabbing any longer.
Introduction
This indicator uses machine learning (Artificial Intelligence) to solve a real human problem.
The Exponential Moving Average ( EMA ) is one of the most used indicators on the planet, yet no one really knows what pair of exponential moving average lengths works best in combination with each other.
A reason for this is because no two EMA lengths are always going to be the best on every instrument, time-frame, and at any given point in time.
The "Intelligent Exponential Moving Average" solves the moving average problem by adapting the period length to match the most profitable combination of exponential moving averages in real time.
How does the Intelligent Exponential Moving Average work?
The artificial intelligence that operates these moving average lengths was created by an algorithm that tests every single combination across the entire chart history of an instrument for maximum profitability in real-time.
No matter what happens, the combination of these exponential moving averages will be the most profitable.
Can we learn from the Intelligent Moving Average?
There are many lessons to be learned from the Intelligent EMA . Most will come with time as it is still a new concept. Adopting the usefulness of this AI will change how we perceive moving averages to work.
Limitations
Ultimately, there are no limiting factors within the range of combinations that has been programmed. The exponential moving averages will operate normally, but may change lengths in unexpected ways - maybe it knows something we don't?
Thresholds
The range of exponential moving average lengths is between 5 to 40. Additional coverage resulted in TradingView server errors.
Future Updates!
Soon, I will be publishing tools to test the AI and visualise what moving average combination the AI is currently using.
Based on the strategy run of the most profitable combo that is in this video, 177 trades were made for under half a percent of profit. For most traders, commissions would have made the strategy a loser at that point. I do understand that this is an indicator, not a strategy. However, if you're using the strategy tester to prove the "most profitable" combinations as the backbone of the indicator, what's the merit in using the indicator if the whole point is to trade with it? In other words, if the "most profitable" combo in this video yielded under .5% over 177 trades, I'm not quite grasping why the indicator is so much more valuable than a standard fast/slow MA combo. Is the idea here to simply give a trader the most profitable set of MA's as part of their toolkit, rather than just using random fast/slow MA's? I guess I could see some merit in that? But trading based solely off the buy and sell signals of this indicator's crossovers seems like a losing game. Not because there is anything wrong with the script, moreso because solely using MA crossovers to trade just doesn't work. If it did, we'd all be rich by buying and selling the crossovers.
You seem to be pretty passionate about the project, that is cool. Looking forward to seeing some live trading and/or backtesting if you continue to pour your heart and soul into it.
GrantPeace
⋅
@Brad.rinker.58, If you are not a fan of the moving average cross-over, then you wouldn't use this script. To be fair, the moving average produces far greater outcomes than the Exponential moving average so you might want to look at this video
As far as commissions go, external factors that influence cost of trading are not considered ans it is not practical to assume a standard cost as it varies so drastically. This is an algorithm that performs one function.
I have developed strategies that consider commissions, liquidity, and latency outside of tradingview, however these are considered strategies and not indicators.
Strategies and indicators just fall under completely different categories because of the holistic approach that is required to make a strategy actually work.
Longshot520
⋅
@GrantPeace, I wasn't knocking the MA crossover as a tool, more as a singular strategy. I really do like the idea of having the most profitable combo of MA's on the chart. While I personally don't use MA crossovers, I think it's foolish not have any sort of MA's on a chart, as they have time and time again shown to be points of interests in price action. Again, my question was geared more towards the use of it as an indicator vs. a strategy. I think the last line you said summed it up perfectly.
zAngus
⋅
Still needs buy and sell indicator as the lines are so close and some backtesting just to see how efficient it indeed is. Great concept still.
ernie76
⋅
excellent!!!
ashrafj1
⋅
great Job keep updating
orangequant1
⋅
I think I said it before but it bears repeating- this is one of the most brilliant approaches to algo theory I've ever seen. Thanks again!