Neurax. Pro 5
2009 Solo Productions
Introducing. Neurax Pro 5
Multiple regression and neural network systems usually look at individual horses . one at a time.
In contrast, Neurax Pro 5 uses proprietary Simulated Annealing Technology. in a more global approach where the race is analyzed as a whole taking into account all horse interactions simultaneously. Thus, Neurax Pro 5 provides a level of analysis that other handicapping products cannot.
Simply put, Neurax Pro 5 searches for the .most likely. finish order of a race. The software always outputs the first five predicted horses in finish position order. An odds line is added to aid betting.
Neurax Pro 5 learns .by example. via a training process whereby previous races and their outcomes are presented to the software so that appropriate .rules. and probabilities (i.e., .costs.) can be discovered and used for later live handicapping. The entire race is analyzed with all horse interactions occurring at once . just like in a real horse race! The software seeks to discover the most likely finish order for the first five horses of any race.
The software comes with one professionally trained system (SIM) that .simulates. the outcome of any race. By providing your own data, you can train more systems (i.e., create a new SIM) for different tracks, different race classes, race distances, etc. You can even tell Neurax Pro 5 to use a different SIM for different types of races or track conditions.
Simulated Annealing is a general approach to problem solving that can be implemented in various ways. The basic idea is to look at various sequences or permutations . for example, a proposed finish order . and to systematically vary this sequence in a manner that eventually delivers the one sequence with lowest .cost. (or highest probability). Sometimes simulated annealing uses statistical methods but sometimes it does not. As a general approach, it strives to find more .global. solutions to a problem . as opposed to more .local. solutions other approaches may give.
Simulated annealing has its historic origins in molecular physics but it has long been known to be a useful learning algorithm. Until recently, how to use simulated annealing approaches in horse handicapping has remained a mystery.
Neurax Pro 5 requires at least a 3GHz Pentium and 20 MB of free disk space. Any version of Windows XP, Windows 2000, Windows 2003 or Windows Vista is fine. A dual-core Pentium (HT) is useful but not necessary. It is feasible to have a slower laptop that you take to the racetrack for handicapping only; perform training and profit/loss analysis on your faster computer (3GHz+) at your home or office.
Sharing SIM Models
User trained SIM models are stored in the SIM folder. A single file (e.g., mysim.SIM) contains all model information, so moving or copying this file to another computer (put it into the SIM folder on the target machine) has the effect of .transplanting. the SIM model. Thus, you can share trained SIM models with your friends.
Transparent Unzipping
Neurax Pro 5 operates on two types of data files: Horse data files (NRP) which contain information about races that have not yet run (i.e., Daily Racing Form and Track Program information), and Exotic Results data files (XRD) which contain results data for races that have previously run.
NRP and XRD files may be downloaded from Bloodstock Research (BRIS) via Internet (brisnet.com). Usually these files are compressed to save space and lessen transmission times. Compressed files (ZIP files) are then saved on your computer. Uncompressing (i.e., unzipping) these files will produce the corresponding NRP or XRD file.
Optionally, users may choose to leave these files compressed. In this case, ZIP files encountered by Neurax Pro 5 will be automatically unzipped as needed.
Neurax Pro 5 treats all ZIP files encountered as equivalent to the corresponding NRP or XRD file that would result from unzipping. All ZIP files are displayed as though they have already been unzipped. When a compressed NRP or XRD file is actually needed by Neurax Pro 5, unzipping is performed "on the fly" in a manner transparent to the user.
The decision whether to unzip files after download or to leave them compressed is left up to the user. Compressed files take approximately 1/4 the disk space of their unzipped equivalents. Conversely, because compressed files must be unzipped by Neurax Pro 5 before they can be used, performance (of some tasks) will be somewhat slower.
Handicapping Races with Neurax Pro 5
During handicapping, predicted race results are generated using Neurax Pro 5's advanced Simulated Annealing Technology ..
After downloading the Horse (NRP) data files from BRIS for the racetracks and dates desired (using communications software of your choice to any folder you like), select Handicap Races... under the Handicap menu. Simply find the downloaded Horse data files (NRP+XRD) on your computer and add the race cards you want to the handicap list. Select a SIM (simulated annealing model) to use, and press Handicap.
The Find NRP Files section allows you to navigate through your file system, adding files to be handicapped to the handicap list. The NRP Files listbox contains a list of all NRP (Horse data files) in the current folder. Double-clicking on a folder name goes into that folder. Pressing the UpOneLevel button goes up one level in the directory structure. Selecting a different drive letter goes to the root of that drive. Once an NRP file is located, to add it to the handicap list, simply select it (click on it) and press Add File (alternatively, you may simply double-click on the filename). Pressing the Search button will search the selected drive, setting the current folder to that containing the most recently downloaded Horse (NRP) data file.this is useful if the location of recently downloaded files is not known (or you simply forgot).
The Handicap List section maintains a list of files to be handicapped (i.e., the handicap list). This list is "sticky" in that all files placed in the list remain in the list until the user removes them (or exits Neurax Pro 5); this allows easy application of different SIM models to the same set of files. Selecting a file and pressing the Remove button will remove it from the handicap list (the file itself is untouched). Pressing the Clear List button will remove all filenames from the handicap list. If you make a mistake while editing the handicap list, pressing Cancel will undo the most recent changes. Selecting a file (in the handicap list) and pressing the Edit Scratches button invokes the Scratch Editor, which allows editing of late scratches for a given card (the NRP file itself is modified).
One of several SIM models may be used for handicapping, depending on user preference. In Single SIM mode, all data files in the handicap list will be handicapped using the chosen SIM.
Pressing the Handicap button will handicap all files in the handicap list.
Handicap results are displayed on-screen with one window per track card. Maneuvering between windows may be done by clicking on the window title bar or by selecting the window at the bottom of the Handicap menu. At any time, resizing the main window (by maximizing/restoring it or by stretching one of its borders) will cause all handicap result windows to be resized (to fit).. The PgUp, PgDown, and arrow keys may be used to aid viewing.
.By default, the main window is maximized. To change the default to a standard size window, just restore the main window. Conversely, to change the default back to maximizing the main window, simply maximize the main window.
The Scratch Editor (invoked by selecting a file in the handicap list and pressing the Edit Scratches button) can be used to remove horses from consideration that have scratched.this actually permanently edits the Horse (NRP) data file itself. This is the preferred method of dealing with late scratches since a new finish order and odds line will be generated to reflect the new field (during handicapping); since horse interactions are very important in a simulated annealing approach, using the scratch editor is strongly recommended.
If a horse scratches at the last minute (and you have no computer access), simply cross it off the list (of predicted finish order) and continue as if it were never there; if this method is used, the odds line and score values will be somewhat off since all contenders are used in their calculation.predicted finish order usually will remain mostly unaffected; if more than one horse (listed in the top five) scratches skip the race entirely.
After Neurax Pro 5 determines the most likely finish order for a given race (using our Simulated Annealing Technology), a score is assigned to each individual horse that roughly indicates the relative odds of that horse winning (among its competitors)*. The score is used to calculate an odds-line (i.e., odds to win for each horse). A horse's Wodds. value designates the calculated win odds for that horse.horses that run at odds significantly greater than this may signal lucrative betting opportunities. For example, if a horse is predicted to have a 50% chance of winning it's odds (Wodds) are listed as even money (1:1); if a horse is predicted to have a 25% chance of winning it's odds are listed as 3:1. In other words, the odds listed are the computed "break even" odds for a given horse.
*score = estimated win probability * 100.
.A horse's Wodds value is simply calculated from it's score: Wodds=[1/(score/100)]-1.
Horse data files downloaded close to race time will have track program information in them, eliminating the need to download or purchase a separate track program. Program numbers, morning line odds, and allowed wagering (e.g., QUINELLA, etc.) are included whenever present.
Glossary of terms always present:
PP post position
Score Horse's overall score (0-100). Score values are basically its win probability * 100.
Wodds Odds of winning (to $1) [for overlay/underlay detection]
Rx Today's medication (L=Lasix, B=Bute)
Age horse's age and sex (f=filly, m=mare, c=colt, h=horse, g=gelding)
Wgt today's weight carried by this horse
Glossary of terms present if track program information is available:
Pgm Program #. Where the program # and the post position are identical, only one number appears.
Modds Morning Line odds (to $1)
Handicap results may be printed to generate hard copy. While viewing any handicap results window, simply pressing the printer icon (on the main menu bar) will print that card. To print all cards or groups of cards, select the Print Handicap Results... option under the Handicap menu.users may specify that all results are to be printed or individually select which cards are of interest. Printer options may be modified via the Printer Setup menu item in the Handicap menu (paper margins are determined programmatically and are not user-selectable).
SIM Mapping
Normally a single SIM is applied to an entire race card. However, advanced users are given the option of specifying which SIM is to be used for a given race type.
A SIM map (or "SI map") consists of a table that associates a SIM with its corresponding race types. During handicapping, each race is handicapped using its corresponding SIM (described in the SI map).
Race types are defined by specifying a race filter criteria consisting of conditions that must be met. Conditions may include the track code*, surface (turf/dirt/synthetic), distance (route/sprint), track condition (wet/dry). and class (e.g., Stakes) for a race. If a race meets all conditions within a criteria, the associated SIM is used. If a race does not meet the criteria (i.e., one or more conditions are not satisfied), the next criterion is checked (in order) until a matching criterion is found or the SI map is exhausted. If a race matches none of the criteria in the SI map, it is skipped (i.e., not handicapped).
*Track code where the race was actually run (e.g., the Simulcast track if it was a simulcast) is always used during all filtering operations.
.This allows different SIM models to be used to generate WET and DRY predictions. Training separate SIM models for WET and DRY conditions may make some sense. Normally a single SIM is used to generate a single prediction for WET and DRY track conditions (i.e., they are the same prediction).
To view/edit SI maps, select Handicap Races... from the Handicap menu and then press the Edit SI Maps button. Within each SI map, a comments field is provided as an organization aid.enter a description of the map in any format you wish (it is for display purposes only). Individual criterion lines may be enabled/disabled individually. Pressing Reset Map resets a map to default values. Pressing Cancel discards any changes made (to all SI maps).
DataSets
A dataset is a special database file created from an arbitrary collection of Horse race data files (NRP) and their corresponding exotic results files (XRD). Once created, datasets can be used to train new SIM models and to generate Profit/Loss analyses to test SIM performance.
However, before a dataset can be used it must be created. The dataset creation process simply consists of specifying a folder that contains the NRP and XRD files to be included in the dataset*. In essence, data files (NRP & XRD) are processed, "producing" a dataset. Once created, datasets may be renamed and deleted as needed.
*Both the NRP and associated XRD files must reside in the same folder. Once a dataset is created, the files that were used in its creation (NRP/XRD) may be deleted, moved, etc. since the dataset is a file separate from those used to create it.
To create a dataset, place the NRP files and their associated XRD files in a folder by themselves and select Create DataSet... under the Database menu. Locate the folder containing the data files to use, select a name for your dataset* and press Build. The Append option allows files to be added to the end of an existing dataset.
*Long filenames are used throughout (dataset names and SIM names).
The NRP+XRD Data File Location section allows you to navigate through your file system to locate the folder containing the data files that are to be included in the dataset. Double-clicking on a folder name goes into that folder. Pressing the UpOneLevel button goes up one level in the directory structure. Selecting a different drive letter goes to the root of that drive.
All NRP and XRD files in the current folder are automatically displayed in the Matched Pairs section. Un-paired files (an NRP file without results or vise-versa) are displayed (in the Left over files section), so you can download the missing files from BRIS if desired. Un-paired files cannot be included in a dataset, since necessary information is missing.
Existing datasets may be renamed by selecting the Rename DataSet... menu item.
Existing datasets may be deleted by selecting the Delete DataSet... menu item. Select the dataset(s) to be deleted; holding the Ctrl key down will enable multiple selection (dragging can be done to select ranges). Pressing OK will delete all selected datasets.
Selecting Split DataSet... allows for splitting of an original (source) dataset into a train dataset (used for training SIM models) and a test dataset (used to gauge generalization of trained SIM models). A specified percentage of the original (source) dataset is copied into the (newly created) test dataset, with the remainder going into the (newly created) train dataset. The original (source) dataset is left unchanged (i.e., race data is copied).
By selecting Combine DataSets... multiple datasets may be combined (concatenated) to produce a new dataset. Select the datasets that are to be combined; holding the Ctrl key down will enable multiple selections (dragging can be done to select ranges)*. (Selecting only a single dataset has the effect of making a duplicate.) Enter a name for the new dataset. Pressing OK will begin the process.
*Holding the Ctrl key down and mouse clicking selects/unselects another dataset. Holding the Ctrl key down, pressing the mouse button (and holding) while dragging will select/unselect ranges.
Three dataset summary reports are available. An Exotics Report... generates mean and median exotic payouts (for $2 bet) by track; checking the Copy to Clipboard option, generates a comma delimited version of the report and puts it into the Windows clipboard for easy pasting into spreadsheets, etc. Payouts vary widely among tracks.it seems prudent not to waste one's time betting exotics at tracks where the likely payoffs are small. A Daily Double Report. generates mean and median data for Daily Double, Pick Three, Pick Four and Pick Six payouts by track; checking the Copy to Clipboard option puts a comma delimited version of the report into the Windows clipboard. The DataSet Contents Report... shows dataset contents by card; checking the Copy to Clipboard option here copies the report text (in human readable format) into the Windows clipboard for easy pasting into a word processor, etc.
Profit/Loss Analysis
Performance of any SIM may be tested on arbitrary data sets, producing a Profit/Loss analysis for different betting strategies..
Select Profit/Loss Analysis... from the Database menu to test a given SIM on a previously created dataset. You will be asked to choose (from a list) which SIM you wish to use and also (from another list) which dataset you would like to test on.
Races within a dataset may be filtered by race class., distance., track condition, surface, and track initials (BRIS track code)*. Only races within the dataset that match the filter criteria will be used during testing (i.e., profit/loss analysis).
*Profit/Loss filter criteria are "sticky" in that they remain as set until the Reset button is pressed or the program is exited.this facilitates enumeration. Routes are races of 7.5 furlongs or longer.
.Selecting "Custom" enables custom criteria setup, where any number of specific criteria may be enabled (only races that match one or more of the checked criteria are included in the profit/loss analysis).
Testing a dataset produces a profit/loss analysis for different bet strategies. This profit/loss analysis may be viewed on screen and/or printed. PgUp, PgDown and the Arrow keys may be used to aid on-screen viewing; also, clicking on a given line will display it in blue, acting as a bookmark, to aid viewing.
Normally, all odds values indicate actual track odds (i.e., the odds to win at race start). However if the Use Morning Line Odds option is enabled, all betting strategies relying on odds (for bet placement) use morning line odds instead of actual track odds. (e.g., WIN_2TO1 places a win bet if the morning line odds for the top pick are 2:1 or greater).
.If no morning line odds are available for this horse, the bet is assumed not placed.
Enabling the Short Report option generates a profit/loss analysis report without an extended odds breakdown.
Enabling the Copy to Clipboard option places a comma delimited text version of the profit/loss analysis report in the Windows clipboard, allowing for easy pasting into a spreadsheet or word processor.
· Note that some races in a dataset may be skipped, because complete data is not available.
Betting Strategies (used in Profit/Loss Analysis)
This section contains a list of all bet strategies used during a Profit/Loss Analysis and their corresponding meaning.
In the following discussion: the A horse refers to the predicted winner; the B horse refers to the predicted place horse; the C horse refers to the predicted show horse.
WIN
· Bet A horse (predicted winner) to win.
WIN_SCOREss
· Bet A horse to win, only if that horse's score value is ss or more.
WIN_OVERLAY
· Bet A horse to win, only if track win odds for that horse exceed calculated win odds (Wodds value). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_OVERLAYn
· Bet A horse to win, only if track win odds for that horse exceed n times calculated win odds. i.e., bet is placed only when odds>(n*Wodds). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_QFAV
· Bet A horse to win, only if it is a quasi-favorite (i.e., track win odds for that horse are 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_nTO1
· Bet A horse to win, only if track win odds for that horse are n:1 or higher. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_ssP_QFAV
· Bet A horse to win, only if that horse's score is ss or more and it is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_ssP_nTO1
· Bet A horse to win, only if that horse's score is ss or more and it's track win odds are n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
WIN_TOPn
· All top n horses bet to win.
PLACE
· Bet A horse (predicted winner) to place.
PLACE_SCOREss
· Bet A horse to place, only if that horse's score is ss or more.
PLACE_QFAV
· Bet A horse to place, only if that horse is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
PLACE_nTO1
· Bet A horse to place, only if that horse's track win odds are n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
PLACE_ssP_QFAV
· Bet A horse to place, only if that horse's score is ss or more and it is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
PLACE_ssP_nTO1
· Bet A horse to place, only if that horse's score is ss or more and it's track win odds are n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
SHOW
· Bet A horse (predicted winner) to show.
SHOW_SCOREss
· Bet A horse to show, only if that horse's score is ss or more.
SHOW_QFAV
· Bet A horse to show, only if that horse is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
SHOW_nTO1
· Bet A horse to show, only if that horse's track win odds are n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
SHOW_ssP_QFAV
· Bet A horse to show, only if that horse's score is ss or more and it is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
SHOW_ssP_nTO1
· Bet A horse to show, only if that horse's score is ss or more and it's track win odds are n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BWIN
· Bet B horse (predicted place horse) to win.
BWIN_OVERLAY
· Bet B horse to win, only if track win odds for that horse exceed calculated win odds (Wodds value). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BWIN_OVERLAYn
· Bet B horse to win, only if track win odds for that horse exceed n times calculated win odds. i.e., bet is placed only when odds>(n*Wodds). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BPLACE
· Bet B horse to place.
BSHOW
· Bet B horse to show.
CWIN
· Bet C horse (predicted show horse) to win.
CWIN_OVERLAY
· Bet C horse to win, only if track win odds for that horse exceed calculated win odds (Wodds value). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
CWIN_OVERLAYn
· Bet C horse to win, only if track win odds for that horse exceed n times calculated win odds. i.e., bet is placed only when odds>(n*Wodds). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
CPLACE
· Bet C horse to place.
CSHOW
· Bet C horse to show.
TOP2WIN
· Bet both the A and B horse to win.
TOP2WIN_SCOREss
· Bet both the A and B horse to win, only if the A horse score is ss or more.
TOP2WIN_BOTHss
· Bet both the A and B horse to win, only if both the A horse score and B horse score is ss or more.
TOP2WIN_nTO1
· Bet both the A and B horse to win, only if the average win odds [(oddsA+oddsB)/2] is n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
TOP2PLACE
· Bet both the A and B horse to place.
TOP2PLACE_SCOREss
· Bet both the A and B horse to place, only if the A horse score is ss or more.
TOP2PLACE_BOTHss
· Bet both the A and B horse to place, only if both the A horse score and B horse score is ss or more.
TOP2SHOW
· Bet both the A and B horse to show.
ACROSS_BOARD
· Bet A horse to win, place and show.
ACROSS_BOARD_QFAV
· Bet A horse to win, place and show, only if it is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
PLACE_N_SHOW
· Bet A horse to place and show.
PLACE_N_SHOW_QFAV
· Bet A horse to place and show, only if it is a quasi-favorite (track win odds 2:1 or less). If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
B_ACROSS_BOARD
· Bet B horse to win, place and show.
C_ACROSS_BOARD
· Bet C horse to win, place and show.
BC_WIN
· Bet both the B and C horse to win.
BC_PLACE
· Bet both the B and C horse to place.
BC_SHOW
· Bet both the B and C horse to show.
EXACTA
· Straight exacta bet (A horse followed by B horse).
· Straight exacta bet (AB), only if both the A horse score and B horse score is ss or more.
EXACTA_SCOREss
· Straight exacta bet (AB), only if the A horse score is ss or more.
EXACTA_KEYn
· Straight exacta bet, where the A horse is "keyed" with the next n horses. For this bet to pay, the A horse must win with any one of the next n horses completing the ticket. {note: n bets are placed.the A horse is paired with each of the next n horses.}
EXACTA_nTO1
· Straight exacta bet, only if the average win odds [(oddsA+oddsB)/2] is n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BOX2EXACTA
· Two exacta bets (AB,BA).
BOX2EXACTA_BOTHss
· Two exacta bets (AB,BA), only if both the A horse score and the B horse score is ss or more.
· Two exacta bets (AB,BA), only if the A horse score is ss or more.
BOX2EXACTA_nTO1
· Two exacta bets (AB,BA), only if the average win odds [(oddsA+oddsB)/2] is n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
QUINELLA
· Quinella bet (AB).
QUINELLA_BOTHss
· Quinella bet, only if both the A horse score and the B horse score is ss or more.
QUINELLA_SCOREss
· Quinella bet, only if the A horse score is ss or more.
QUINELLA_KEYn
· Quinella bet, where the A horse is "keyed" with the next n horses. For this bet to pay, the A horse must win (or place) with any one of the next n horses completing the ticket. {note: n bets are placed.the A horse is paired with each of the next n horses.}
QUINELLA_nTO1
· Quinella bet, only if the average win odds [(oddsA+oddsB)/2] is n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BOX3EXACTA
· Three horse box exacta.
BOX3EXACTA_SCOREss
· Three horse box exacta, only if the A horse score is ss or more.
BOX3EXACTA_SUMnn
· Three horse box exacta, only if the sum of first three (A,B,C) horse scores is nn or more.
TRIFECTA_KEYn
· Trifecta bet, where the A horse is "keyed" with the next n horses. For this bet to pay, the A horse must win with any two of the next n horses completing the ticket. {note: n!/(n-2)! bets are placed.all combinations of the next n horses filling place and show positions.}
BOX3TRIFECTA
· Three horse box trifecta.
BOX3TRIFECTA_nTO1
· Three horse box trifecta bet, only if the average win odds [(oddsA+oddsB+oddsC)/3] is n:1 or more. If the Use Morning Line Odds option is enabled, morning line odds will be used in lieu of actual track odds.
BOX3TRIFECTA_SUMnn
· Three horse box trifecta, only if the sum of first three (A,B,C) horse scores is nn or more.
BOX4TRIFECTA
· Four horse box trifecta.
SUPERFECTA_KEYn
· Superfecta bet, where the A horse is "keyed" with the next n horses. For this bet to pay, the A horse must win with any three of the next n horses completing the ticket. {note: n!/(n-3)! bets are placed.all combinations of the next n horses filling place, show and fourth positions.}
BOX3SUPERFECTA
· Superfecta bet boxing the top 3 horses with the predicted fourth place horse. For this bet to pay, the top 3 horses must finish 1,2,3 (in any order) and the predicted fourth place horse must finish fourth.
BOX4SUPERFECTA
· Four horse box superfecta.
Training SIM models
Users may train additional SIM models using a dataset of their choosing. Training occurs in two steps. To train a new SIM, select the Prep SIM... option under the Train menu. Select the dataset you wish to use for training (previously created) and enter a name for your new SIM*. Pressing the Prep button will begin preparations for the training process; to perform actual training select the Train SIM. option (also under the Train menu).
*Long filenames are used throughout (dataset names and SIM names).
Training is asynchronous . it occurs in the background while you do other things! Use the .Stop Training. menu item (in the Train menu) if you wish to stop training before it completes normally. Training progress is displayed in the title bar.
The .Prep SIM.. option prepares the data file to be used during training (it.s a large file . usually 20 to 40 GB) and the .Train SIM.. option performs the actual training. Additionally, the .Train SIM.. option may be selected after training to .train again. . i.e., to make additional training passes (the filter criteria and dataset used as well as the target SIM name are the same as last setup via .Prep SIM..).
Races within a dataset may be filtered by race class, distance, track condition, surface, and track initials (BRIS track code). Only races within the dataset that match the filter criteria will be used for training.
Training is particularly useful for training SIM models that are sensitive to individual track differences.
During normal training, Neurax Pro 5 starts from a random starting point and builds a new SIM model that converges onto the given dataset. This is .from scratch. training and requires adequate data! The learning pressure controls how far to push learning (if you are training from a random starting point use .high. or .very high.). Learning pressure is additive: performing two Train SIM phases is roughly equivalent to one Train SIM phase with double the learning pressure.
Alternatively, training may begin from an existing SIM. This feature may be useful for "fine tuning" or "updating" a SIM model with new data (contained within a dataset). When applied to an existing SIM, learning pressure may be viewed as a continuum determining the degree of adaptation to new data that the existing SIM model will undergo.
For best results, it is strongly recommended that at least 500 hundred or 1,000 races be used for training. If you use fewer races for training, it is recommended that only the "low" learning pressure setting be used. With very small amounts of data, "over-fitting" can become a problem which results in degraded SIM model performance.
After training, it is recommended that a Profit/Loss Analysis be done using the newly trained SIM model on the same dataset used for training so that one can get a feel for the effects of training*. Experimentation is recommended to determine which learning pressure setting gives the best results. The time it takes for training is a factor here since as learning pressure increases one level the total training time approximately doubles.
*Since the dataset was used for training, results can be expected to be somewhat higher than on novel data. Ideally, an additional test dataset (not used during training) should be tested (i.e., Profit/Loss Analysis) to gauge how well results will generalize to novel data.
For the really serious horseplayers who wish to train their own custom SIM models "from scratch" (as opposed to "fine tuning" an existing SIM model): starting from a random starting point using a "very high" (.high. is usually okay too) learning pressure with a training set of at least 1,000 races should give the very best results. Since training on large datasets takes hours (or even days), simply letting the software run overnight (etc.) may be easiest; if you don.t have the time you may train via doing a single prep (Prep SIM) with .low. learning pressure and then select .Train SIM. multiple times (depending on your schedule).
In general, using more training data precludes the danger of "over-fitting" and increases "generalization" to novel data. Our experiments indicate that the optimal number of races to use for training lies somewhere within the 1,000 to 3,000 range; exceeding 3,000 races seems to offer little additional benefit; using less than 1,000 races for .from scratch. training is strongly discouraged!
One may also create SIM models that are optimized for specific bet performance. First, start with an existing SIM or from a random starting point. Then, repeatedly perform training using low (or medium) learning pressure five to ten times (or more if you like), each time using the most recent SIM as the next starting point . or you can just repeatedly choose .Train SIM.. after data has been prepared (via .Prep SIM..) and do a profit/loss analysis after each training to gauge progress. Finally, use the SIM that maximizes bet performance on an independent test set; delete the other (temporary) SIM models; the .Train SIM.. option if used repeatedly saves each training to the (same) target SIM as specified in the .Prep SIM.. stage (so no deleting is necessary). For example: divide a dataset into train and test datasets (via Split DataSet); starting from a random starting point, train (using the train dataset) producing SIM 'A1'; starting from SIM 'A1' train again producing SIM 'A2'; starting from SIM 'A2' train again producing SIM 'A3', etc; perform profit/loss analysis (using test dataset) on SIM models 'A1'...AN, picking the one that maximizes bet performance.
Two test sets are recommended so that .generalization. may be more adequately measured . ideally there should be little variation of performance between two independent test sets (which are both independent from the training dataset).
It is probably easier to perform one .Prep SIM.. phase followed by multiple .Train SIM.. passes while doing a profit/loss analysis after each training; you may rename the target SIM after each training phase (the next training phase via Train SIM will continue training from its last point . it does not use the target SIM but keeps a temporary SIM internally).
If learning is pushed too far, over-fitting may occur which is recognizable by a decrease in performance on an independent test set (i.e., on data not used for training) . however, the algorithms used by Neurax Pro 5 make over-fitting a remote possibility only, but one you should keep aware of; if performance drops considerably after a given training phase (Train SIM.) then over-fitting has probably occurred . just back up to the last completed training phase (or add more data). Over-fitting is usually the result of too little training data; if you have adequate training data (1,000 races or more) then over-fitting is highly unlikely.
We offer only suggestions here. There are too many possible training strategies to comprehensively list here. Users are urged to experiment with all available training options to find what best suits their needs.
Track Codes
The following is a list of the most common North American Track Codes:
AKS Ak-Sar-Ben APX Arlington Park
AQU Aqueduct ATL Atlantic City
BEL Belmont Park BEU Beulah Park
BGD Blue Grass Downs BIR Birmingham Race Course
BMF Bay Meadows Fair BMX Bay Meadows
CBY Canterbury Park CDX Churchill Downs
CNL Colonial Downs CRC Calder Race Course
CTX Charles Town DEL Delaware Park
DET Detroit Race Course DMR Del Mar
DUE Dueling Grounds ELP Ellis Park
EMD Emerald Downs EVD Evangeline Downs
FER Ferndale FEX Fort Erie
FGX Fair Grounds FLX Finger Lakes
FMP Fairmount Park FNO Fresno
FPX Fairplex FSX Flagstaff
GGX Golden Gate GPX Gulfstream Park
GSX Garden State Park HAW Hawthorne
HIA Hialean Park HOL Hollywood Park
HOO Hoosier Park HOU Sam Houston Race Park
HST Hastings Park KEE Keeneland
LAD Louisiana Downs LAX Los Alamitos
LRL Laurel Race Course LSX Lone Star
MED Meadowlands MNR Mountaineer Park
MTH Monmouth Park OPX Oaklawn Park
PEN Penn National PHA Philadelphia Park
PIM Pimlico PLA Playfair
PLN Pleasanton PMX Portland Meadows
PRE Prescott Downs PRM Prairie Meadows
RDX River Downs RET Retama Race Park
RKM Rockingham Park RPX Remington Park
SAC Sacramento SAR Saratoga
SAX Santa Anita SOL Solano
SPT Sportsman's Park SRX Santa Rosa
STK Stockton SUF Suffolk Downs
TAM Tampa Bay Downs TDN Thistledown
TIM Timonium TPX Turfway Park
TRM Trinity Meadows TUP Turf Paradise
WDS Woodlands WOX Woodbine
YMX Yakima Meadows
Disclaimer:
Past performance is not a guarantee of future success. Individual results may vary. The user assumes the responsibility of verifying all mathematical assumptions used. Solo Productions is in no way responsible for any losses incurred by the use of its Neurax Pro 5 product.
Acknowledgements:
Neurax Pro 5 uses special compression and de-compression utilities from the Info-ZIP group (zip.exe and unzip.exe) to process ZIP files. Info-ZIP's software (Zip, UnZip and related utilities) is free and can be obtained as source code or executables from Internet/WWW sites, including http://www.info-zip.org/pub/infozip/; see infoziplicense.txt (shipped with this product) or ftp://ftp.info-zip.org/pub/infozip/license.html for Info-ZIP.s licensing agreement.