ANN Trend PredictionThis trend indicator utilizes an artificial neural network (ANN) to predict the next market reversal within a certain range of previous candles. The larger the range of previous candles you set, the fewer reversals will be predicted, and trends will tend to last longer.
The ANN is trained on the BTCUSD 4-hour chart, so using it on other assets or timeframes may yield suboptimal results. It takes three input values: the closing price, the Stochastic RSI, and a Choppiness Indicator. Based on these inputs, the ANN categorizes the current candle as part of an uptrend, downtrend, or as undefined.
Compared to an EMA-based trend indicator, this ANN identifies reversals several candles earlier. It achieves this by detecting subtle patterns in the input values that typically appear before a market turnaround. These patterns are somewhat specific to that chosen asset and timeframe.
The results are displayed using rows of triangles that indicate the predicted price direction. The price levels of the triangles correspond to the closing price at the last reversal. The area between the triangle row and the price is colored green if the ANN correctly predicted the move, and red if it did not.
This indicator is designed to showcase the capabilities and potential of ANNs, and is not intended for actual trading use. The ANN can be trained on any other input values, assets and timeframes for several predictions tasks.
You can use the Predicted_Trend_Signal of this Indicator in any backtest indicator. In the Backtester just grap the Predicted_Trend_Signal. downtrend = 1, uptrend = -1, undefined = 0
Feel free to write me a comment.
Ann
Machine Learning: Perceptron-based strategyPerceptron-based strategy
Description:
The Learning Perceptron is the simplest possible artificial neural network (ANN), consisting of just a single neuron and capable of learning a certain class of binary classification problems. The idea behind ANNs is that by selecting good values for the weight parameters (and the bias), the ANN can model the relationships between the inputs and some target.
Generally, ANN neurons receive a number of inputs, weight each of those inputs, sum the weights, and then transform that sum using a special function called an activation function. The output of that activation function is then either used as the prediction (in a single neuron model) or is combined with the outputs of other neurons for further use in more complex models.
The purpose of the activation function is to take the input signal (that’s the weighted sum of the inputs and the bias) and turn it into an output signal. Think of this activation function as firing (activating) the neuron when it returns 1, and doing nothing when it returns 0. This sort of computation is accomplished with a function called step function: f(z) = {1 if z > 0 else 0}. This function then transforms any weighted sum of the inputs and converts it into a binary output (either 1 or 0). The trick to making this useful is finding (learning) a set of weights that lead to good predictions using this activation function.
Training our perceptron is simply a matter of initializing the weights to zero (or random value) and then implementing the perceptron learning rule, which just updates the weights based on the error of each observation with the current weights. This has the effect of moving the classifier’s decision boundary in the direction that would have helped it classify the last observation correctly. This is achieved via a for loop which iterates over each observation, making a prediction of each observation, calculating the error of that prediction and then updating the weights accordingly. In this way, weights are gradually updated until they converge. Each sweep through the training data is called an epoch.
In this script the perceptron is retrained on each new bar trying to classify this bar by drawing the moving average curve above or below the bar.
This script was tested with BTCUSD, USDJPY, and EURUSD.
Note: TradingViews's playback feature helps to see this strategy in action.
Warning: Signals ARE repainting.
Style tags: Trend Following, Trend Analysis
Asset class: Equities, Futures, ETFs, Currencies and Commodities
Dataset: FX Minutes/Hours+/Days
Candlesticks ANN for Stock Markets TF : 1WHello, this script consists of training candlesticks with Artificial Neural Networks (ANN).
In addition to the first series, candlesticks' bodies and wicks were also introduced as training inputs.
The inputs are individually trained to find the relationship between the subsequent historical value of all candlestick values 1.(High,Low,Close,Open)
The outputs are adapted to the current values with a simple forecast code.
Once the OHLC value is found, the exponential moving averages of 5 and 20 periods are used.
Reminder : OHLC = (Open + High + Close + Low ) / 4
First version :
Script is designed for S&P 500 Indices,Funds,ETFs, especially S&P 500 Stocks,and for all liquid Stocks all around the World.
NOTE: This script is only suitable for 1W time-frame for Stocks.
The average training error rates are less than 5 per thousand for each candlestick variable. (Average Error < 0.005 )
I've just finished it and haven't tested it in detail.
So let's use it carefully as a supporter.
Best regards !
ANN Next Coming Candlestick Forecast SPX 1D v1.0WARNING:
Experimental and incomplete.
Script is open to development and will be developed.
This is just version 1.0
STRUCTURE
This script is trained according to the open, close, high and low values of the bars.
It is tried to predict the future values of opening, closing, high and low values.
A few simple codes were used to correlate expectation with current values. (You can see between line 129 - 159 )
Therefore, they are all individually trained.
You can see in functions.
The average training error of each variable is less than 0.011.
NOTE :
This script is designed for experimental use on S & P 500 and connected instruments only on 1-day bars.
The Plotcandle function is inspired by the following script of alexgrover :
Since we estimate the next values, our error rates should be much lower for all candlestick values. This is just first version to show logic.
I will continue to look for other variables to reach average error = 0.001 - 0.005 for each candlestick status.
Feel free to use and improve , this is open-source.
Best regards.
ANN BTC MTF Golden Cross Period MACDHi, this is the MACD version of the ANN BTC Multi Timeframe Script.
The MACD Periods were approximated to the Golden Cross values.
MACD Lengths :
Signal Length = 25
Fast Length = 50
Slow Length = 200
Regards.
ANN BTC MTF CM Sling Shot SystemHi all, this script was created as a result of ANN training in all time frames of bitcoin data.
Trained data is built on Chris Moody's Sling Shot system.
CM Sling Shot System :
This system automatically generates the ANN output for all time periods.
Therefore, it has multi-time-frame feature.
Artificial Neural Networks training details:
Average Errors
1 minute = 0.005570
3 minutes = 0.006674
5 minutes = 0.007067
15 minutes = 0.010000
30 minutes = 0.009398
45 minutes = 0.010000
1 Hour = 0.006848
2 Hours = 0.006901
3 Hours = 0.009608
4 Hours = 0.009774
1 Day = 0.010000
1 Week = 0.010000
The results look good (All Average Error <= 0.01 ), the Sling Shot Method is also good, but you can also refer to historically slower period averages to filter these arrows a bit more. I leave the decision to you.
Best regards.
Macroeconomic Artificial Neural Networks
This script was created by training 20 selected macroeconomic data to construct artificial neural networks on the S&P 500 index.
No technical analysis data were used.
The average error rate is 0.01.
In this respect, there is a strong relationship between the index and macroeconomic data.
Although it affects the whole world,I personally recommend using it under the following conditions: S&P 500 and related ETFs in 1W time-frame (TF = 1W SPX500USD, SP1!, SPY, SPX etc. )
Macroeconomic Parameters
Effective Federal Funds Rate (FEDFUNDS)
Initial Claims (ICSA)
Civilian Unemployment Rate (UNRATE)
10 Year Treasury Constant Maturity Rate (DGS10)
Gross Domestic Product , 1 Decimal (GDP)
Trade Weighted US Dollar Index : Major Currencies (DTWEXM)
Consumer Price Index For All Urban Consumers (CPIAUCSL)
M1 Money Stock (M1)
M2 Money Stock (M2)
2 - Year Treasury Constant Maturity Rate (DGS2)
30 Year Treasury Constant Maturity Rate (DGS30)
Industrial Production Index (INDPRO)
5-Year Treasury Constant Maturity Rate (FRED : DGS5)
Light Weight Vehicle Sales: Autos and Light Trucks (ALTSALES)
Civilian Employment Population Ratio (EMRATIO)
Capacity Utilization (TOTAL INDUSTRY) (TCU)
Average (Mean) Duration Of Unemployment (UEMPMEAN)
Manufacturing Employment Index (MAN_EMPL)
Manufacturers' New Orders (NEWORDER)
ISM Manufacturing Index (MAN : PMI)
Artificial Neural Network (ANN) Training Details :
Learning cycles: 16231
AutoSave cycles: 100
Grid
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 998
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Network
Input nodes connected: 19
Hidden layer 1 nodes: 2
Hidden layer 2 nodes: 0
Hidden layer 3 nodes: 0
Output nodes: 1
Controls
Learning rate: 0.1000
Momentum: 0.8000 (Optimized)
Target error: 0.0100
Training error: 0.010000
NOTE : Alerts added . The red histogram represents the bear market and the green histogram represents the bull market.
Bars subject to region changes are shown as background colors. (Teal = Bull , Maroon = Bear Market )
I hope it will be useful in your studies and analysis, regards.
Blockchain Artificial Neural NetworksI found a very high correlation in a research-based Artificial Neural Networks.(ANN)
Trained only on daily bars with blockchain data and Bitcoin closing price.
NOTE: It does not repaint strictly during the weekly time frame. (TF = 1W)
Use only for Bitcoin .
Blockchain data can be repainted in the daily time zone according to the description time.
Alarms are available.
And you can also paint bar colors from the menu by region.
After making reminders, let's share the details of this interesting research:
INPUTS :
1. Average Block Size
2. Api Blockchain Size
3. Miners Revenue
4. Hash Rate
5. Bitcoin Cost Per Transaction
6. Bitcoin USD Exchange Trade Volume
7. Bitcoin Total Number of Transactions
OUTPUTS :
1. One day next price close (Historical)
TRAINING DETAILS :
Learning cycles: 1096436
AutoSave cycles: 100
Grid :
Input columns: 7
Output columns: 1
Excluded columns: 0
Training example rows: 446
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Network :
Input nodes connected: 7
Hidden layer 1 nodes: 5
Hidden layer 2 nodes: 0
Hidden layer 3 nodes: 0
Output nodes: 1
Controls :
Learning rate: 0.1000
Momentum: 0.8000
Target error: 0.0100
Training error: 0.010571
The average training error is really low, almost worth the target.
Without using technical analysis data, we established Artificial Neural Networks with blockchain data.
Interesting!
NAND PerceptronExperimental NAND Perceptron based upon Python template that aims to predict NAND Gate Outputs. A Perceptron is one of the foundational building blocks of nearly all advanced Neural Network layers and models for Algo trading and Machine Learning.
The goal behind this script was threefold:
To prove and demonstrate that an ACTUAL working neural net can be implemented in Pine, even if incomplete.
To pave the way for other traders and coders to iterate on this script and push the boundaries of Tradingview strategies and indicators.
To see if a self-contained neural network component for parameter optimization within Pinescript was hypothetically possible.
NOTE: This is a highly experimental proof of concept - this is NOT a ready-made template to include or integrate into existing strategies and indicators, yet (emphasis YET - neural networks have a lot of potential utility and potential when utilized and implemented properly).
Hardcoded NAND Gate outputs with Bias column (X0):
// NAND Gate + X0 Bias and Y-true
// X0 // X1 // X2 // Y
// 1 // 0 // 0 // 1
// 1 // 0 // 1 // 1
// 1 // 1 // 0 // 1
// 1 // 1 // 1 // 0
Column X0 is bias feature/input
Column X1 and X2 are the NAND Gate
Column Y is the y-true values for the NAND gate
yhat is the prediction at that timestep
F0,F1,F2,F3 are the Dot products of the Weights (W0,W1,W2) and the input features (X0,X1,X2)
Learning rate and activation function threshold are enabled by default as input parameters
Uncomment sections for more training iterations/epochs:
Loop optimizations would be amazing to have for a selectable length for training iterations/epochs but I'm not sure if it's possible in Pine with how this script is structured.
Error metrics and loss have not been implemented due to difficulty with script length and iterations vs epochs - I haven't been able to configure the input parameters to successfully predict the right values for all four y-true values for the NAND gate (only been able to get 3/4; If you're able to get all four predictions to be correct, let me know, please).
// //---- REFERENCE for final output
// A3 := 1, y0 true
// B3 := 1, y1 true
// C3 := 1, y2 true
// D3 := 0, y3 true
PLEASE READ: Source article/template and main code reference:
towardsdatascience.com
towardsdatascience.com
towardsdatascience.com
ANN GOLD WORLDWIDE This script consists of converting the value of 1 gram and / or 1 ounce of gold according to the national currencies into a system with artificial neural networks.
Why did I feel such a need?
Even though the printed products in the market are digitally circulated, only precious metals are available in full or near full.
Silver is difficult to carry because you have to buy too much because the unit price is low.
Platinum is very difficult to find and used in industry.
Gold is both practical and has less volatile movements, even more balanced than dollars, to preserve the value of money.
Uncertainty and tensions benefit gold.
Obviously this is my own opinion and is not worth the investment advice:
If there is to be an economic crisis, it is obvious that the dollar will rise against the emerging currencies, but I expect a crisis where gold and the dollar will rise together.
The world has been on a mercantilist line more than ever!
Spot gold can be bought from goldsmiths and banks.
I think this command will benefit people everywhere but in economies that are subject to developing currencies.
Now we can look at the details:
All you have to do is load the appropriate chart and select it from the menu.
Thus, the system will adjust itself to that instrument.
MENU and Tickers :
"GOLD" : XAUUSD or GC1! or GOLD (Average error = 0.0128)
"GOLDSILVER" : XAUXAG or GOLDSILVER (Gold Silver Ratio ) ( Average error : 0.01 )
"GOLD CZK " : XAUUSD/USDCZK ( 1 Ounce Gold Czech Koruna) ( Average error = 0.010879 )
"GOLD NZD " : XAUUSD/USDNZD ( 1 Ounce Gold New Zealand Dollar ) (Average error = 0.010736 )
"GOLD EURO" : XAUUSD/USDEUR ( 1 Ounce Gold Euro) ( Average error = 0.010000 )
"GOLD HUF " : XAUUSD/USDHUF ( 1 Ounce Gold Hungarian Forint ) ( Average error = 0.010000 )
"GOLD INR " : XAUUSD/USDINR (1 Ounce Gold Indian Rupee ) (Average error = 0.010458 )
"GOLD DKK" : XAUUSD/USDDKK (1 Ounce Gold Danish Krone) (Average error = 0.010671 )
"GOLD CHF" : XAUUSD/USDCHF (1 Ounce Gold Swiss Franc ) (Average error = 0.010967 )
"GOLD CNH" : XAUUSD/USDCNH(1 Ounce Gold Offshore RMB) (Average error = 0.012017 )
"GOLD MXN" : XAUUSD/USDMXN(1 Ounce Gold Mexican Peso) (Average error = 0.010000 )
"GOLD PLN" : XAUUSD/USDPLN (1 Ounce Gold Polish Zloty ) (Average error = 0.010173 )
"GOLD ZAR" : XAUUSD/USDZAR (1 Ounce Gold South African Rand (Average error = 0.010484 )
"GOLD NOK" : XAUUSD/USDNOK (1 Ounce Gold Norwegian Krone ) (Average error = 0.010842 )
"GOLD TRY" : XAUUSD/USDTRY (1 Ounce Gold Turkish Lira ) (Average error = 0.010000 )
"GOLD THB" : XAUUSD/USDTHB (1 Ounce Gold Thai Baht ) (Average error = 0.011747 )
Important note : XAUUSD/USDCUR = 1 Ounce Gold , XAUUSD/31.1*USDCUR = 1 gram Gold (CUR = Currency )
If you want to physically hold it, look gram value, because as far as I know, all goldsmiths and jewelleries in the world are selling gram gold.
I think that this command is the most useful and the concrete one that I have ever written.
I end my sentences with this anonymous proverb :
"Even if gold falls into the mud, it's still gold ! "
ANN MACD : 25 IN 1 SCRIPTIn this script, I tried to fit deep learning series to 1 command system up to the maximum point.
After selecting the ticker, select the instrument from the menu and the system will automatically turn on the appropriate ann system.
Listed instruments with alternative tickers and error rates:
WTI : West Texas Intermediate (WTICOUSD , USOIL , CL1! ) Average error : 0.007593
BRENT : Brent Crude Oil (BCOUSD , UKOIL , BB1! ) Average error : 0.006591
GOLD : XAUUSD , GOLD , GC1! Average error : 0.012767
SP500 : S&P 500 Index (SPX500USD , SP1!) Average error : 0.011650
EURUSD : Eurodollar (EURUSD , 6E1! , FCEU1!) Average error : 0.005500
ETHUSD : Ethereum (ETHUSD , ETHUSDT ) Average error : 0.009378
BTCUSD : Bitcoin (BTCUSD , BTCUSDT , XBTUSD , BTC1!) Average error : 0.01050
GBPUSD : British Pound (GBPUSD,6B1! , GBP1!) Average error : 0.009999
USDJPY : US Dollar / Japanese Yen (USDJPY , FCUY1!) Average error : 0.009198
USDCHF : US Dollar / Swiss Franc (USDCHF , FCUF1! ) Average error : 0.009999
USDCAD : Us Dollar / Canadian Dollar (USDCAD) Average error : 0.012162
SOYBNUSD : Soybean (SOYBNUSD , ZS1!) Average error : 0.010000
CORNUSD : Corn (ZC1! ) Average error : 0.007574
NATGASUSD : Natural Gas (NATGASUSD , NG1!) Average error : 0.010000
SUGARUSD : Sugar (SUGARUSD , SB1! ) Average error : 0.011081
WHEATUSD : Wheat (WHEATUSD , ZW1!) Average error : 0.009980
XPTUSD : Platinum (XPTUSD , PL1! ) Average error : 0.009964
XU030 : Borsa Istanbul 30 Futures ( XU030 , XU030D1! ) Average error : 0.010727
VIX : S & P 500 Volatility Index (VX1! , VIX ) Average error : 0.009999
YM : E - Mini Dow Futures (YM1! ) Average error : 0.010819
ES : S&P 500 E-Mini Futures (ES1! ) Average error : 0.010709
GAZP : Gazprom Futures (GAZP , GZ1! ) Average error : 0.008442
SSE : Shangai Stock Exchange Composite (Index ) ( 000001 ) Average error : 0.011287
XRPUSD : Ripple (XRPUSD , XRPUSDT ) Average error : 0.009803
Note 1 : Australian Dollar (AUDUSD , AUD1! , FCAU1! ) : Instrument has been removed because it has an average error rate of over 0.13.
The average error rate is 0.1850.
I didn't delete it from the menu just because there was so much request,
You can use.
Note 2 : Friends have too many requests, it took me a week in total and 1 other script that I'll share in 2 days.
Reaching these error rates is a very difficult task, and when I keep at a low learning rate, they are trained for a very long time.
If I don't see the error rate at an average low, I increase the layers and go back into a longer process.
It takes me 45 minutes per instrument to command artificial neural networks, so I'll release one more open source, and then we'll be laying 70-80 percent of the world trade volume with artificial neural networks.
Note 3 :
I would like to thank wroclai for helping me with this script.
This script is subject to MIT License on behalf of both of us.
You can review my original idea scripts from my Github page.
You can use it free but if you are going to modify it, just quote this script .
I hope it will help everyone, after 1-2 days I will share another ann script that I think is of the same importance as this, stay tuned.
Regards , Noldo .
ANN MACD WTI (West Texas Intermediate) This script created by training WTI 4 hour data , 7 indicators and 12 Guppy Exponential Moving Averages.
Details :
Learning cycles: 1
AutoSave cycles: 100
Training error: 0.007593 ( Smaller than average target ! )
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 300
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 2
Hidden layer 2 nodes: 6
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.7000
Momentum: 0.8000
Target error: 0.0100
Special thanks to wroclai for his great effort.
Deep learning series will continue. But I need to rest my eyes a little :)
Stay tuned ! Regards.
ANN MACD BRENT CRUDE OIL (UKOIL) This script trained with Brent Crude Oil data including 7 basic indicators and 12 Guppy Exponential Moving Averages .
Details :
Learning cycles: 1
Training error: 0.006591 ( Smaller than 0.01 ! )
AutoSave cycles: 100
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 300
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 2
Hidden layer 2 nodes: 6
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.7000
Momentum: 0.8000
Target error: 0.0100
Note : Alerts added .
Special thanks to wroclai for his great effort.
Deep learning series will continue , stay tuned ! Regards.
ANN MACD GOLD (XAUUSD)This script aims to establish artificial neural networks with gold data.(4H)
Details :
Learning cycles: 329818
Training error: 0.012767 ( Slightly above average but negligible.)
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 300
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 5
Hidden layer 2 nodes: 1
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.7000
Momentum: 0.8000
Target error: 0.0100
NOTE : Alarms added.
And special thanks to dear wroclai for his great effort.
Deep learning series will continue . Stay tuned! Regards.
ANN MACD S&P 500 This script is formed by training the S & P 500 Index with various indicators. Details :
Learning cycles: 78089
AutoSave cycles: 100
Training error: 0.011650 (Far less than the target, but acceptable.)
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 300
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 2
Hidden layer 2 nodes: 1
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.7000
Momentum: 0.8000
Target error: 0.0100
Note : Thanks for dear wroclai for his great effort .
Deep learning series will continue . Stay tuned! Regards.
ANN MACD EURUSD (FX) Hello , this script is trained with eurusd 4-hour data. (550 columns) Details :
Learning cycles: 8327
AutoSave cycles: 100
Training error: 0.005500 ( That's a very good error coefficient.)
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 550
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 2
Hidden layer 2 nodes: 5
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.6000
Momentum: 0.8000
Target error: 0.0055
NOTE : Use with EURUSD only.
Alarms added.
Thanks dear wroclai for his great effort.
Deep learning series will continue ! Stay tuned.
Regards , Noldo .
ANN MACD ETHEREUM
This script is trained with Ethereum (Timeframe : 4 hours ).
Details :
Input columns: 19
Output columns: 1
Excluded columns: 0
Training example rows: 300
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 19
Hidden layer 1 nodes: 8
Hidden layer 2 nodes: 1
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate: 0.7000
Momentum: 0.8000
Training error: 0.009378 ( That's a very good error coefficient. )
Many thanks to wroclai for help.
Deep learning series will continue!
ANN MACD BTC v2.0 This script is the 2nd version of the BTC Deep Learning (ANN) system.
Created with the following indicators and tools:
RSI
MACD
MOM
Bollinger Bands
Guppy Exponential Moving Averages:
(3,5,8,10,12,15,30,35,40,45,50,60)
Note: I was inspired by the CM Guppy Ema script.
Thank you very much to dear wroclai for his great help.
He has been a big help in the deep learning series.
That's why the licenses in this series are for both of us.
I'm sharing these series and thats the first. Stay tuned and regards!
Note : Alerts added.
SPY FRACTAL S-R LEVELS (FIXED ANN MACD)
This is a fractal version of my deep learning script for SPY
In addition, buy and sell conditions may appear in bar colors in green and red.
You can choose from the menu if you wish.
Fractal codes do not belong to me.
So I didn't put any license.
You can use it as you want, you can change and modify.
Regards.Noldo
BTC FRACTAL ANN S-R LEVELS (Fixed ANN MACD)
This script is an adaptation of my deep learning system for Bitcoin to fractals.
Fractal codes are not belong to me. Original :
The code for the Deep learning (ANN MACD BTC) work belongs to me. Original:
I didn't get license for this script because the fractal codes don't belong to me.You can use it for any purpose.
This command can be a very helpful guide.You can use that fractals with your indicators for Bitcoin.
You can also combine these levels with ANN - MACD - BTC script.
Scripts about Artificial Neural Networks (ANN) will continue soon !
I hope it will help us to gain insight into technical analysis.
Best regards. Noldo.
ANN MACD (BTC)
Logic is correct.
But I prefer to say experimental because the sample set is narrow. (300 columns)
Let's start:
6 inputs : Volume Change , Bollinger Low Band chg. , Bollinger Mid Band chg., Bollinger Up Band chg. , RSI change , MACD histogram change.
1 output : Future bar change (Historical)
Training timeframe : 15 mins (Analysis TF > 4 hours (My opinion))
Learning cycles : 337
Training error: 0.009999
Input columns: 6
Output columns: 1
Excluded columns: 0
Grid
Training example rows: 301
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Network
Input nodes connected: 6
Hidden layer 1 nodes: 8
Hidden layer 2 nodes: 0
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate : 0.6 Momentum : 0.8
More info :
EDIT : This code is open source under the MIT License. If you have any improvements or corrections to suggest, please send me a pull request via the github repository github.com
ANN MACD Future Forecast (SPY 1D) NOTE : Deep learning was conducted in a narrow sample set for testing purposes. So this script is Experimental .
This system is based on the following article and is inspired by an external program:
hackernoon.com
None of the artificial neural networks in Tradingview work and are not based on completely correct logic. Unlike others in this system:
IMPORTANT NOTE: If the tangent activation function is used, the input data must also have tangent values (compared to the previous values of 1 bar).
Inputs were prepared according to this judgment.
1. The tangent function which is the activation function is written correctly. (The tangent function in the article: ActivationFunctionTanh (v) => (1 - exp (-2 * v)) / (1 + exp (-2 * v)))
2. Missing bias parts in the formulas were added.
3. The output function is taken from the next day (historical), so that the next bar can be predicted, which is the truth.
4.The forecast value of the next bar is subtracted from the current bar change and the market direction is determined.
5.When the future forecast and the current close are added together, the resulting data is called seed.
The seed carries data both from the present and from yesterday and from the future.
6.And this seed was subjected to the MACD method.
Thus, due to exponential averages, more importance will be given to recent developments and
The acceleration situations will show us the direction.
However, a short position should be taken for crossover and a long position for crossunder .
Because the predicted values work in reverse.Even though we use the same period (9,12,26) it is much faster!
7. There is no future code that can cause Repaint.
However, the color after closing should be checked.
The system is completely correct.
However, a very narrow sample was selected.
100 data: Tangent diffs ; volume change, bollinger bands values changes (Upband , Midband , Lowband) and LazyBear's Squeeze Momentum Indicator (SQZMOM_LB) change and the next bar data (historical) price change were put into the deep learning test.
IMPORTANT NOTE : The larger the sample set and the more effective dependent variables, the higher the hit rate of the deep learning test!
EDIT : This code is open source under the MIT License. If you have any improvements or corrections to suggest, please send me a pull request via the github repository github.com
Stay tuned. Best regards!
ANN [Original Version Repainting Fixed]It's the ANN, the original version with repainting issue got taken care of. The result as you can see, it's might not be that good compare to original. If anyone want to put it into a strategy, feel free. I do not recommend you to use this as a stand alone. Trade with your own risk.
Maybe trade the opposite way might do it lol
