Set, "Max Bars In Chart" parameter - minimum at 2400.

In New Version 3.3, please, set "Mx Bars In Chart" parameter - minimum at 18000.

Recommended Settings:

Recommended Settings 2 (Simple):

   Manual Instruction:

    Before Using The Indicator For The First Time Or After A Long Break,

    It Is Best To Preload The Necessary Data.

    To Do This, First, In The Terminal Settings, 

    Set The Value Of Maximum Number Of Bars In Chart To At Least 17116.

    Then Restart The Terminal To Apply The New Settings.

    After That, Open The Chart Window Of The Required Financial Instrument,

    Switch To Month Timeframe, Switch To Minimum Scale Of Chart

    And Wait Until The Entire History Is Loaded.

    The Next Step Is A Gradual Switch From The Monthly Timeframe

    To The Lower One Up To 1 Minute.

    Now Everything Is Ready For The First Launch Of The Indicator.

    4 Artificial Neural Network Ensembles Predict:

    1: 60 Bars (about 1 Minute)

    Note: For Efficient Training Of This Neural Network Ensemble

    And Obtaining Adequate Predictions, A Minimum Of

    1 Week Of Non-Stop Training Is Required.

    2: 300 Bars (about 5 Minutes)

    Note: For Efficient Training Of This Neural Network Ensemble

    And Obtaining Adequate Predictions, A Minimum Of

    1 Month Of Non-Stop Training Is Required.

    3: 1800 Bars (about 30 Minutes)

    Note: For Efficient Training Of This Neural Network Ensemble

    And Obtaining Adequate Predictions, A Minimum Of

    0.5 Year Of Non-Stop Training Is Required.

    4: 14400 Bars (about 240 Minutes)

    Note: For Efficient Training Of This Neural Network Ensemble

    And Obtaining Adequate Predictions, A Minimum Of

    2.5 Year Of Non-Stop Training Is Required.

    To Reset Weights, Please, Delete Appropriate Global Variables

    In Terminal Settings:

    0) 0+Prefix+Delete_Me_To_Reset_Network_ - To Reset Slave Network Weights Of 0 Ensemble

    1) 1+Prefix+Delete_Me_To_Reset_Network_ - To Reset Master Network Weights Of 0 Ensemble

    2) 2+Prefix+Delete_Me_To_Reset_Network_ - To Reset Slave Network Weights Of 1 Ensemble

    3) 3+Prefix+Delete_Me_To_Reset_Network_ - To Reset Master Network Weights Of 1 Ensemble

    4) 4+Prefix+Delete_Me_To_Reset_Network_ - To Reset Slave Network Weights Of 2 Ensemble

    5) 5+Prefix+Delete_Me_To_Reset_Network_ - To Reset Master Network Weights Of 2 Ensemble

    6) 6+Prefix+Delete_Me_To_Reset_Network_ - To Reset Slave Network Weights Of 3 Ensemble

    7) 7+Prefix+Delete_Me_To_Reset_Network_ - To Reset Master Network Weights Of 3 Ensemble

    There Are 32 Simple Settings:

    0) General Color Scheme:

      a) _black_inf     - Dark Informative Theme 

      b) _black_smt     - Dark Simplified Theme 

      c) _white_inf     - Light Informative Theme 

      d) _white_smt     - Light Simplified Theme

      Note: Neural Networks Do Not Work In Simplified Mode

    1) General Calc Type:

      a) _01_sec        - 1 Second TimeFrame

      b) _02_sec        - 2 Second TimeFrame

      c) _12_sec        - 12 Second TimeFrame

      d) _cur_tm        - Chart TimeFrame

      Note: Neural Networks Do Not Work In Other Than _01_sec Mode

    2) General Data Type:

      a) _normalzd      - Normalized Data From 1 To -1

      b) _standard      - Data Displayed In Same Format As On The Main Chart

      Note: Neural Networks Do Not Work In _standard Mode

    3) Neural Networks Ensemble Use:

      a) false          - Not Use Neural Networks Ensemble

      b) true           - Use Neural Networks Ensemble

      Note: Before Enabling This Setting, Make Sure

      Your Computer Meets The Minimum Requirements:

      Processor Has More Than 10 Cores,

      Free RAM Is More Than 16 GB.

      Neural Networks Will Not Work In Modes:

      General Color Scheme: _black_smt Or _white_smt,

      General Calc Type: _02_sec, _12_sec Or _cur_tm,

      General Data Type: _standard,

      Additional Info Show: false.

    4) Neural Networks Save Input Data:

      a) false          - Save Neural Networks Input Data

      b) true           - Not Save Neural Networks Input Data

      Note: This Will Be Useful If You Often Change The Settings Of The Terminal,

            Chart Or Indicator, But Do Not Want To Waste Valuable Time

            Spent On Accumulating Input Data For Neural Networks.

            But This Only Applies To Slave Networks Ofg Ensambles.

    5) Neural Networks Learning Passes Per Second: 

      a) 20

      Note: It Depends On The Processing Power Of Your Computer.

      For Example: If You Have A 5 GHz Modern Processor,

      Then You Can Set The Value Up To 100,

      If 2.5 GHz Old Processor, Then It Is Better To Lower The Values.

      The Minimum Value Is 2.

    6) Neural Networks S0 Mem Lim 59.04 Mb Of RAM Per Hour:

      a) 120

      Note: It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    7) Neural Networks M0 Mem Lim 29.56 Mb Of RAM Per Hour:

      a) 48

      Note: This Is Important To Set This Value Up To

      3 Or 4 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    8) Neural Networks S1 Mem Lim 294.97 Of RAM Mb Per Hour:

      a) 120

      Note: It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    9) Neural Networks M1 Mem Lim 29.56 Of RAM Mb Per Hour:

      a) 48

      Note: This Is Important To Set This Value Up To

      3 Or 4 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    10) Neural Networks S2 Mem Lim 294.97 Mb Of RAM Per Hour:

      a) 120

      Note: It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    11) Neural Networks M2 Mem Lim 29.56 Mb Of RAM Per Hour:

      a) 48

      Note: This Is Important To Set This Value Up To

      3 Or 4 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    12) Neural Networks S3 Mem Lim 2.87 Mb Of RAM Per Day:

      a) 120

      Note: It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    13) Neural Networks M3 Mem Lim 11.8 Mb Of RAM Per Day:

      a) 48

      Note: This Is Important To Set This Value Up To

      3 Or 4 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of RAM In Your Computer.

      Find Out The Amount Of Free RAM, And Set The Appropriate Value.

    14) Neural Networks Ensemble 0 Save 59.04 Mb Of HD Per Hour:

      a) 100

      Note: This Is Important To Set This Value Up To

      0.5 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of Hard Drive Size In Your Computer.

      Find Out The Amount Of Free Hard Drive Size, And Set The Appropriate Value.

    15) Neural Networks Ensemble 1 Save 294.97 Mb Of HD Per Hour:

      a) 100

      Note: This Is Important To Set This Value Up To

      0.5 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of Hard Drive Size In Your Computer.

      Find Out The Amount Of Free Hard Drive Size, And Set The Appropriate Value.

    16) Neural Networks Ensemble 2 Save 294.97 Mb Of HD Per Hour:

      a) 100

      Note: This Is Important To Set This Value Up To

      0.5 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of Hard Drive Size In Your Computer.

      Find Out The Amount Of Free Hard Drive Size, And Set The Appropriate Value.

    17) Neural Networks Ensemble 3 Save 2.87 Mb Of HD Per Day:

      a) 100

      Note: This Is Important To Set This Value Up To

      0.5 Times Less Than Appropriate Slave Network.

      It Depends On The Amount Of Hard Drive Size In Your Computer.

      Find Out The Amount Of Free Hard Drive Size, And Set The Appropriate Value.

    18) Additional Info Show:

      a) false          - Not Show Additional Info

      b) true           - Show Additional Info

      Note: Neural Networks Do Not Work In Not Show Additional Info Mode

    19) Additional Info Subwindow:

      a) 0

      Note: It Depends On The Configuration Of Your Workspace

      And The Use Of Other Indicators.

      The Main Chart Window Is 0.

    20) Additional Info Grid Size:

      a) _3X2

      b) _3X1

      c) _2X1

      Note: It Depends On The Configuration Of Your Workspace.

    21) Additional Info Grid Start:

      a) _M             - Start From Minute TimeFrame

      b) _H             - Start From Hour TimeFrame

      c) _W             - Start From Week TimeFrame

      Note: It Depends On The Configuration Of Your Workspace.

      This Will Be Useful If You Use Additional Info Grid Size

      Smaller Than 3X2.

    22) Additional Info Cell 0 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    23) Additional Info Cell 1 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    24) Additional Info Cell 2 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    25) Additional Info Cell 3 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    26) Additional Info Cell 4 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    27) Additional Info Cell 5 (General/Uniform):

      a) false          - Unifrom Type Is The Same As Standard Japanese Candles

      b) true           - Generalized Type iS Compressed format of standard Japanese candles.

      Note: Generalized Type Will Be Useful If You Use Small Display

      Or Simplification Of Perception Of Information And More Concentration Into Work.

    28) Additional Info Objects Name:

      a) "_SuperFilter_"- Additional Info Objects Name Prefix

      Note: This Will Be Useful If You Use Additional Programms On Same Chart

      And Don't Want Names Of Objects To Overlap With Each Other.

    29) Additional Info Arrow Size:

      a) _SmallToLarge  - Weak Signal Has Small Graphic Object Size

      b) _LargeToSmall  - Weak Signal Has Large Graphic Object Size

      c) _SameLarge     - All Signal Graphic Object Have Same Large Size

      d) _SameSmall     - All Signal Graphic Object Have Same Small Size

    30) Additional Info Voice Notifications:

      a) false          - Not Use Additional Info Voice Notifications

      b) true           - Use Additional Info Voice Notifications

    31) Additional Info Voice Notifications Risk Limit:

      a) _Lowest        - Additional Info Voice Notifications Is Activated Only On Week Timeframe

      b) _Low           - Additional Info Voice Notifications Is Activated From Day Timeframe

      c) _Medium        - Additional Info Voice Notifications Is Activated From H4 Timeframe

      d) _High          - Additional Info Voice Notifications Is Activated From M30 Timeframe

      e) _Highest       - Additional Info Voice Notifications Is Activated From M5 Timeframe

If you want to use this indicator in conjunction with an expert,

then use the corresponding global variables published by this indicator,

which display the predictions of the corresponding ensembles of neural networks.