Appearance
Source files: 17 | Classes: 152 | Methods: 16 | Enums: 5
GTOS.Intelligence.Algorithms.Audio
InstrumentClassificationResult
struct
Source: IntelligenceAlgorithmsAudio.cs
Constants and Fields
Class
InstrumentClass
Confidence
float
Features
TimbreFeatures
Methods
DetectPitch
PitchDetectionResult DetectPitch ( GTAudioBuffer audio, int startFrame, int windowSize, PitchDetectionMethod method )
IntelligenceAlgorithmsAudio
static class
Intelligence algorithms for audio analysis and recognition.
Used by Music and AudioProcessing networks for AI-driven workflows.
Source: IntelligenceAlgorithmsAudio.cs
Enumerations
PitchDetectionMethod
Values: Autocorrelation, YIN, PYIN, HPS, Cepstrum, InvalidParameter
Constants and Fields
CalculationFailure
const float
OnsetDetectionResult
struct
Source: IntelligenceAlgorithmsAudio.cs
Constants and Fields
Duration
float
OffsetFrame
int
OnsetFrame
int
Strength
float
PercussiveClassificationResult
struct
Source: IntelligenceAlgorithmsAudio.cs
Constants and Fields
Confidence
float
SpectralCentroid
float
TransientFrame
int
Type
PercussiveType
PitchDetectionResult
struct
Source: IntelligenceAlgorithmsAudio.cs
Constants and Fields
Confidence
float
FrameIndex
int
FrequencyHz
float
IsVoiced
bool
TimbreFeatures
struct
Source: IntelligenceAlgorithmsAudio.cs
Constants and Fields
MFCC
float[]
MFCCCount
int
RMS
float
SpectralCentroid
float
SpectralFlatness
float
SpectralFlux
float
SpectralRolloff
float
ZeroCrossingRate
float
GTOS.Intelligence.Networks
IntelligenceAlgorithmsNetworks
static class
Pre-built execution networks for Intelligence Algorithms applications
Source: IntelligenceAlgorithmsNetworks.cs
GTOS.IntelligenceAlgorithms
IntelligenceAlgorithmsExecutionEngine
static class
Intelligence Algorithms Execution Engine
Registers ML/AI domain with Core Execution Engine
Source: IntelligenceAlgorithmsExecutionEngine.cs
GTOS.IntelligenceAlgorithms.AnomalyDetection
AnomalyDetectionML
static class
Anomaly detection intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
AnomalyResult
readonly struct
Anomaly detection result.
WHAT IT CONTAINS:
- Is anomaly flag
- Anomaly score (0-1, higher = more anomalous)
- Type and severity
- Detection confidence
USE THIS FOR:
- Equipment monitoring alerts
- Quality control flagging
- Process deviation detection
- Security breach identification
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
AnomalyScore
readonly double
Confidence
readonly ConfidenceLevel
Index
readonly int
IsAnomaly
readonly bool
Severity
readonly AnomalySeverity
Type
readonly AnomalyType
ChangePointResult
readonly struct
Change point detection result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
ChangePointCount
readonly int
ChangePoints
readonly int[]
ChangeScores
readonly double[]
SignificanceLevel
readonly double
Type
readonly ChangePointType
IsolationResult
readonly struct
Isolation score result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
AveragePathLength
readonly double
IsAnomaly
readonly bool
IsolationScore
readonly double
Severity
readonly AnomalySeverity
LOFResult
readonly struct
Local outlier factor result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
Confidence
readonly ConfidenceLevel
IsOutlier
readonly bool
LocalOutlierFactor
readonly double
NeighborDistances
readonly double[]
Neighbors
readonly int[]
MultivariateAnomalyResult
readonly struct
Multivariate anomaly result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
AnomalousDimensions
readonly int[]
IsAnomaly
readonly bool
MahalanobisDistance
readonly double
Severity
readonly AnomalySeverity
NoveltyResult
readonly struct
Novelty detection result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
Confidence
readonly ConfidenceLevel
DistanceToNormal
readonly double
IsNovel
readonly bool
NoveltyScore
readonly double
OutlierResult
readonly struct
Statistical outlier detection result.
Source: IntelligenceAlgorithmsAnomalyDetection.cs
Constants and Fields
Method
readonly DetectionMethod
OutlierCount
readonly int
OutlierIndices
readonly int[]
OutlierScores
readonly double[]
Threshold
readonly double
GTOS.IntelligenceAlgorithms.Core
AdaptivePrecisionSelector
static class
Context-driven precision selection for adaptive quantization.
WHAT IT DOES:
Automatically determines optimal bit width based on input data
characteristics and expected outcomes for memory and speed efficiency.
TECHNICAL: Context-aware precision selection for BitNet/ByteNet
Source: IntelligenceAlgorithmsCoreAtomics.cs
Common
static class
Pre-built training patterns for common ML scenarios.
SILVIA.IntelligenceAlgorithms.Patterns namespace.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
BinaryClassifier
TrainingResult BinaryClassifier ( ReadOnlySpan<float> trainX, ReadOnlySpan<float> trainY, int inputSize, int hiddenSize = 0, int epochs = 1000 )
Binary Classification - Train a yes/no, pass/fail, accept/reject model.
WHAT IT DOES:
Trains a neural network to make binary decisions (two outcomes only).
Automatically configures a 3-hidden-layer network optimized for binary problems.
INPUTS:
- trainX: Training examples (features) - flatten into 1D array
- trainY: Labels (0 or 1 for each example)
- inputSize: Number of features per example
- hiddenSize: Neurons per hidden layer (0 = auto-calculate)
- epochs: Training iterations (default 1000, increase if accuracy is low)
OUTPUTS:
- TrainingResult with trained model and performance metrics
USE CASES:
- Quality control pass/fail
- Equipment failure prediction (will fail / won't fail)
- Approval decisions (approve / reject)
- Anomaly detection (normal / abnormal)
EXAMPLE:
// Predict equipment failure from 5 sensor readings
float[] sensorData = [...]; // [temp1, pressure1, vibration1, ..., temp2, pressure2, ...]
float[] failures = [0, 0, 1, 0, ...]; // 0=normal, 1=failure
var result = BinaryClassifier(sensorData, failures, inputSize: 5, epochs: 2000);
TRAINING DATA FORMAT:
- trainX: [sample1_feature1, sample1_feature2, ..., sample2_feature1, ...]
- trainY: [sample1_label, sample2_label, ...] where label is 0 or 1
TECHNICAL: 3-layer ReLU network + Sigmoid output, Adam optimizer, Binary Cross-Entropy loss
DataDistribution
readonly struct
Data distribution structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Kurtosis
readonly float
Mean
readonly float
Skewness
readonly float
StandardDeviation
readonly float
DataProcessing
static class
Data preprocessing, normalization, and transformation algorithms.
All methods operate on Span<T> for zero-allocation performance.
No interfaces, no heap allocation, deterministic execution.
IEEE 1990-2024 AI Safety compliant.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
MinMaxNormalize
void MinMaxNormalize ( Span<float> data, float min, float max )
Min-Max Normalization - Scale data to 0-1 range.
WHAT IT DOES:
Rescales all values to fit between 0 and 1, preserving relative distances.
Smallest value becomes 0, largest becomes 1, everything else in between.
INPUTS:
- data: Values to normalize (modified in-place)
- min, max: Original data range
OUTPUTS:
- data: Scaled to [0, 1]
WHEN TO USE:
- Features have different scales (temperature 0-300, pressure 0-100)
- Neural networks (work best with 0-1 inputs)
- Before using distance-based algorithms (KNN, K-Means)
EXAMPLE:
float[] temperatures = [100, 150, 200, 250, 300];
MinMaxNormalize(temperatures, min: 100, max: 300);
// Result: [0, 0.25, 0.5, 0.75, 1.0]
ANALOGY:
Like converting test scores to percentiles. The lowest score becomes 0%,
highest becomes 100%, everyone else falls proportionally in between.
WHY NORMALIZE:
Without it, features with large ranges (1000-10000) dominate
features with small ranges (0-1). Normalization makes them equal.
PROS: Simple, preserves relationships, bounded output
CONS: Sensitive to outliers (one extreme value affects everything)
TECHNICAL: (x - min) / (max - min)
DataRange
readonly struct
Data range structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Max
readonly float
Min
readonly float
Domain
static class
Domain-specific training patterns.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
TimeSeries
TrainingResult TimeSeries ( ReadOnlySpan<float> timeSeriesData, int windowSize, int forecastSteps = 1, int epochs = 1000 )
Time Series Prediction - Forecast future values from historical data.
WHAT IT DOES:
Learns patterns in sequential data to predict what comes next.
Uses a sliding window approach - looks at recent history to forecast future.
INPUTS:
- timeSeriesData: Sequential values (e.g., daily temperatures, hourly sensor readings)
- windowSize: How many past values to look at (e.g., 7 for weekly pattern)
- forecastSteps: How many steps ahead to predict (default 1)
- epochs: Training iterations (default 1000)
OUTPUTS:
- TrainingResult with model that predicts future values from recent history
USE CASES:
- Equipment sensor trend forecasting
- Demand prediction from sales history
- Temperature forecasting from past readings
- Production volume prediction
- Maintenance schedule optimization
EXAMPLE:
// Predict tomorrow's temperature from last 7 days
float[] dailyTemps = [72, 75, 74, 76, 78, 77, 79, 80, 81, ...];
var result = TimeSeries(dailyTemps, windowSize: 7, forecastSteps: 1);
HOW IT WORKS:
windowSize=3, forecastSteps=1, data=[1,2,3,4,5,6]
Training examples:
[1,2,3] ? predict 4
[2,3,4] ? predict 5
[3,4,5] ? predict 6
CHOOSING WINDOW SIZE:
- Daily data with weekly pattern: windowSize = 7
- Hourly data with daily pattern: windowSize = 24
- General rule: windowSize = length of repeating pattern
TECHNICAL: 3-layer Tanh network (good for sequences), Adam, MSE loss
Generic
static class
Generic processing patterns for any domain.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
ProcessData
float[] ProcessData ( ReadOnlySpan<float> data, ModelAsset model )
Process any preprocessed data through model.
IntelligenceConstants
static class
Core constants for intelligence algorithms.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
CalculationFailure
const double
Sentinel value for calculation failures.
MIL-SPEC: Prevents NaN crashes in agentic systems.
LossConfig
readonly struct
Loss function configuration.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Type
readonly LossType
LossFunctions
static class
Loss functions and their gradients for neural network training.
All methods are zero-allocation and use Span<T> for performance.
ISO/IEC 22989:2022 AI Concepts compliant.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
MeanSquaredError
float MeanSquaredError ( ReadOnlySpan<float> predictions, ReadOnlySpan<float> targets )
Mean Squared Error (MSE) - Heavily penalizes big mistakes.
WHAT IT DOES:
Measures how far your predictions are from actual values. Squares the errors,
so being "off by 10" is 100x worse than being "off by 1".
INPUTS:
- predictions: Your model's guesses (e.g., predicted temperatures, yields, costs)
- targets: The actual correct values you're trying to match
OUTPUTS:
- Single number: Average squared error (lower is better, 0 is perfect)
WHEN TO USE:
- Predicting continuous numbers (temperature, pressure, yield, cost, time)
- When large errors are much worse than small errors
- Default choice for regression problems
ANALOGY:
Like grading a test where getting an answer "close" still fails you badly.
Missing a temperature target by 10� counts as 100 times worse than missing by 1�.
AVOID WHEN:
- You have outliers (wild data points) - they get penalized too heavily
- Small and large errors should be treated similarly (use MeanAbsoluteError)
TECHNICAL: (1/n) * S(y_pred - y_true)�
MachineLearning
static class
Traditional machine learning algorithms.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
LinearRegression
void LinearRegression ( ReadOnlySpan<float> x, ReadOnlySpan<float> y, out float slope, out float intercept )
Linear Regression - Fit a straight line to data points.
WHAT IT DOES:
Finds the best-fit line (y = slope*x + intercept) through your data.
The "best" line minimizes the squared distance from all points.
INPUTS:
- x: Independent variable (e.g., temperature, time, input value)
- y: Dependent variable (e.g., yield, cost, output value)
OUTPUTS:
- slope: How much y changes per unit change in x
- intercept: Starting value of y when x=0
USE CASES:
- Temperature vs. yield relationship
- Time vs. cost trend
- Pressure vs. flow rate correlation
- Any two-variable relationship
EXAMPLE:
float[] temperatures = [100, 150, 200, 250, 300];
float[] yields = [85, 90, 95, 97, 99];
LinearRegression(temperatures, yields, out float slope, out float intercept);
// Predict: yield = slope * 175 + intercept
ANALOGY:
Like drawing the best straight line through scattered dots on graph paper.
Some dots above, some below, but the line represents the overall trend.
WHEN TO USE:
- Simple linear relationships
- Quick trend analysis
- Baseline model before trying complex methods
TECHNICAL: Normal Equation w = (X^T * X)^-1 * X^T * y (1D optimized)
MLErrorConstants
static class
Core error constants for ML domain
MIL-SPEC: Standardized error codes for all ML operations
Source: IntelligenceAlgorithmsCoreAtomics.cs
Enumerations
MLCalculationError
Core error codes for all ML calculations
MIL-SPEC: Standardized error handling across all ML subdomains
Values: Success, InvalidMode, InvalidInputData, InvalidContext, CalculationFailure, MemoryAllocationFailure, NetworkExecutionFailure
MLValidationResult
Validation result codes for machine learning operations.
WHAT IT DOES:
Instead of crashing with errors, ML operations return these codes to tell you
exactly what went wrong. Like traffic lights (red/yellow/green) instead of accidents.
WHEN TO USE:
Check these codes after calling validation methods to see if your data is ready.
ANALOGY:
Like a pre-flight checklist - each code tells you which specific check failed
so you can fix it before takeoff (training).
MIL-SPEC: No exceptions, use return codes for deterministic behavior.
Values: Success, InvalidInputShape, InvalidDimensions, InvalidParameters, NullOrEmptyData, InsufficientData, DivisionByZero, NumericalInstability, UnsupportedOperation
NetworkExecutionMode
Core execution modes for all ML networks
MIL-SPEC: Standardized execution modes across all ML subdomains
Values: Training, FineTuning, Inference, Validation, Export
Constants and Fields
BiasesH
float[]
BiasesO
float[]
CALCULATION_FAILURE
const int
CreatedDate
DateTime
DimensionMeans
float[]
DimensionStdDevs
float[]
InputMean
float
InputShape
int[]
InputStdDev
float
INVALID_CONTEXT
const int
INVALID_INPUT
const int
INVALID_MODE
const int
MEMORY_ALLOCATION_FAILURE
const int
ModelType
ModelType
NETWORK_EXECUTION_FAILURE
const int
OutputSize
int
SUCCESS
const int
UsePerDimensionNormalization
bool
Version
int
WeightsHO
float[]
WeightsIH
float[]
Methods
PreprocessInput
void PreprocessInput ( Span<float> input )
Generic input preprocessing - works with any data type.
ModelAssetExtensions
static class
Domain-specific extensions for ModelAsset.
Use these only if you need domain-specific convenience methods.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
PredictFromBitmap
MLValidationResult PredictFromBitmap ( this ModelAsset model, ReadOnlySpan<byte> bitmap, int width, int height, int channels, Span<float> output )
Image processing extension - use if working with bitmap data.
ModelMetadata
struct
Model metadata structure for resource planning.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Enumerations
ModelType
Model type classification - what kind of problem your model solves.
WHAT IT TELLS YOU:
The type of predictions your model makes. Like knowing if a tool is
a hammer, screwdriver, or wrench - tells you what it's designed for.
TYPES EXPLAINED:
- BinaryClassifier: Yes/No decisions (pass/fail, on/off, accept/reject)
- MultiClassClassifier: Choose one from many options (Grade A/B/C/D/F)
- Regression: Predict a number (temperature, cost, yield percentage)
- Autoencoder: Find patterns and compress data (reduce 100 sensors to 10 key factors)
- TimeSeries: Predict future from history (sales forecast, sensor trends)
- Generic: Imported model of unknown type
Values: BinaryClassifier, MultiClassClassifier, Regression, Autoencoder, TimeSeries, Generic
Constants and Fields
InputShape
int[]
ModelSize
long
OutputShape
int[]
RequiredMemory
long
ModelValidator
static class
Model validation and verification algorithms.
SILVIA.IntelligenceAlgorithms.Validation namespace.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
ValidateMetadata
ValidationResult ValidateMetadata ( ModelMetadata metadata )
Validate model metadata.
MIL-SPEC: Returns result code instead of throwing exceptions.
NetworkBuilder
struct
Value-type builder for neural networks. Zero allocation, compile-time safe.
Source: IntelligenceAlgorithmsCoreAtomics.cs
NetworkModeSwitching
static class
Core mode switching logic for all ML networks
MIL-SPEC: Standardized mode switching across all ML subdomains
Source: IntelligenceAlgorithmsCoreAtomics.cs
NeuralNetworks
static class
Neural network operations: activations, layers, pooling, regularization.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
ReLU
void ReLU ( ReadOnlySpan<float> input, Span<float> output )
ReLU (Rectified Linear Unit) - Default activation for 80% of neural networks.
WHAT IT DOES:
Passes positive values unchanged, blocks negative values (sets them to zero).
Simple, fast, and effective.
INPUTS:
- input: Raw neuron outputs (any numbers)
- output: Activated outputs (0 or positive)
WHEN TO USE:
- Default choice for hidden layers
- Fast training needed
- Deep networks (many layers)
ANALOGY:
Like a one-way valve - lets positive signals flow through, blocks negative.
PROS: Fastest activation, simple, works great for most problems
CONS: "Dying ReLU" problem - neurons can get stuck at zero
IF NEURONS DIE: Try LeakyReLU or GELU instead
TECHNICAL: max(0, x)
OnlineLearner
struct
Lightweight learner for real-time updates. Value type, stack-allocated.
Source: IntelligenceAlgorithmsCoreAtomics.cs
OnlineLearning
static class
Real-time learning patterns for online/streaming scenarios.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
Create
OnlineLearner Create ( int inputSize, int outputSize, ActivationType activation = ActivationType.ReLU )
Create online learner for real-time updates.
ONNXAtomics
static class
ONNX Model Integration - Core atomic operations for model import/export.
WHAT IT DOES:
Provides fundamental atomic operations for ONNX model integration,
including parsing, conversion, and optimization for SILVIA execution.
TECHNICAL: Core ONNX operations for all ML algorithms
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
ParseONNXModel
ONNXModelInfo ParseONNXModel ( ReadOnlySpan<byte> onnxData )
Parse ONNX Model Structure - Extract model information from ONNX data.
WHAT IT DOES:
Parses ONNX protobuf structure to extract model metadata,
tensor shapes, and operation definitions.
INPUTS:
- onnxData: Raw ONNX model data (protobuf format)
OUTPUTS:
- Parsed model information
- Tensor shapes and types
- Operation definitions
ALGORITHM:
1. Parse protobuf structure
2. Extract model metadata
3. Parse tensor definitions
4. Extract operation graph
TECHNICAL: ONNX protobuf parsing
ONNXEdge
readonly struct
ONNX edge structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
SourceNode
readonly string
TargetNode
readonly string
TensorName
readonly string
ONNXGraph
readonly struct
ONNX graph structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Edges
readonly ONNXEdge[]
Metadata
readonly ONNXMetadata
Nodes
readonly ONNXNode[]
TensorShapes
readonly int[][]
ONNXMetadata
readonly struct
ONNX metadata structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Description
readonly string
ModelName
readonly string
PropertyKeys
readonly string[]
PropertyValues
readonly string[]
Version
readonly string
ONNXModelInfo
readonly struct
ONNX model information structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
ModelVersion
readonly int
Operations
readonly ONNXOperation[]
ProducerName
readonly string
ProducerVersion
readonly string
TensorShapes
readonly int[][]
TensorTypes
readonly TensorType[]
ONNXModelMetadata
readonly struct
Model metadata structure for ONNX models.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Description
readonly string
ModelName
readonly string
PropertyKeys
readonly string[]
PropertyValues
readonly string[]
Version
readonly string
ONNXNode
readonly struct
ONNX node structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Inputs
readonly string[]
Name
readonly string
OperationType
readonly string
Outputs
readonly string[]
ONNXOperation
readonly struct
ONNX operation structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
AttributeKeys
readonly string[]
AttributeValues
readonly object[]
Name
readonly string
Type
readonly string
ONNXWeights
readonly struct
ONNX weights structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
BiasTensors
readonly float[][]
TensorNames
readonly string[]
WeightTensors
readonly float[][]
Optimization
static class
Optimization algorithms for training neural networks.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
SGD
void SGD ( Span<float> weights, ReadOnlySpan<float> gradients, float learningRate )
Stochastic Gradient Descent (SGD) - Simple, fast, predictable optimizer.
WHAT IT DOES:
Updates model weights by moving them in the direction that reduces error.
Fixed learning rate - you control exactly how fast it learns.
INPUTS:
- weights: Current model parameters (will be modified in-place)
- gradients: Direction to adjust (calculated automatically during training)
- learningRate: How fast to learn (typical: 0.001 to 0.1)
OUTPUTS:
- Updates weights directly (no return value)
WHEN TO USE:
- You want full control over learning speed
- Your problem is well-behaved (smooth error surface)
- Speed is critical (fastest optimizer)
- You have experience tuning learning rates
ANALOGY:
Like driving at constant speed. Simple and predictable, but you need
to manually adjust speed for curves vs. straightaways.
PROS: Fast, simple, predictable, no memory overhead
CONS: Requires manual learning rate tuning, can get stuck
LEARNING RATE GUIDE:
Too high (>0.1): Training unstable, error jumps around
Good (0.001-0.01): Steady progress
Too low (<0.0001): Training too slow
TECHNICAL: w = w - lr * grad
OptimizerConfig
readonly struct
Optimizer configuration structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Beta1
readonly float
Beta2
readonly float
LearningRate
readonly float
Type
readonly OptimizerType
PerformanceResult
readonly struct
Performance benchmarking results.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
AverageInferenceTime
readonly double
ThroughputPerSecond
readonly double
TotalIterations
readonly int
TotalTime
readonly double
QuantizationParameters
readonly struct
Quantization parameters structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
MaxValue
readonly float
Scale
readonly float
Threshold
readonly float
ZeroPoint
readonly float
QuantizedData
readonly struct
Quantized data structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Parameters
readonly QuantizationParameters
Precision
readonly Precision
RandomExtensions
static class
Random number generation extensions for machine learning.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
NextGaussian
double NextGaussian ( this Random random, double mean = 0.0, double stddev = 1.0 )
Generate Gaussian (normal) random number using Box-Muller transform.
SILVIAModel
readonly struct
SILVIA model structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Biases
readonly float[]
Metadata
readonly ONNXModelMetadata
ModelType
readonly ModelType
Precision
readonly Precision
TensorShapes
readonly int[]
Weights
readonly float[]
SpanExtensions
static class
Span extension methods for compatibility.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Methods
ToArray
float[] ToArray ( this ReadOnlySpan<float> span )
Convert ReadOnlySpan to array for compatibility.
TrainingConfig
readonly struct
Training configuration parameters.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
BatchSize
readonly int
Epochs
readonly int
L2Lambda
readonly float
TrainingResult
readonly struct
Training result structure.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
EpochsCompleted
readonly int
FinalLoss
readonly float
ValidationConfig
readonly struct
Validation configuration options.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
FailFast
readonly bool
MaxErrors
readonly int
Tolerance
readonly float
ValidationResult
readonly struct
Validation result structure with error reporting.
MIL-SPEC: Return codes instead of exceptions.
Source: IntelligenceAlgorithmsCoreAtomics.cs
Constants and Fields
Message
readonly string
ResultCode
readonly MLValidationResult
GTOS.IntelligenceAlgorithms.Diffusion
AttentionResult
readonly struct
Attention computation result.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
AttentionScore
readonly float
AttentionWeights
readonly float[]
Output
readonly float[]
Type
readonly AttentionType
CameraParameters
readonly struct
Camera parameters for 3D projection.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
FocalLengthX
readonly float
FocalLengthY
readonly float
PrincipalPointX
readonly float
PrincipalPointY
readonly float
Comprehensive3DResult
readonly struct
Comprehensive 3D reconstruction result.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
NumAngles
readonly int
NumVideoFrames
readonly int
PhysicsSDF
readonly PhysicsSDF
PointCloud
readonly PointCloud3D
Quality
readonly ReconstructionQuality
DiffusionML
static class
Diffusion models intelligence atomic calculations.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsDiffusion.cs
DiffusionResult
readonly struct
Diffusion sampling result with generated content and metadata.
WHAT IT CONTAINS:
- Generated image/audio/text
- Sampling parameters used
- Quality metrics
- Generation time
USE THIS FOR:
- Image generation results
- Quality assessment
- Parameter optimization
- Batch processing
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
Algorithm
readonly SamplingAlgorithm
Channels
readonly int
GeneratedContent
readonly float[]
GenerationTimeMs
readonly double
GuidanceScale
readonly float
Height
readonly int
QualityScore
readonly float
Seed
readonly int
Steps
readonly int
Width
readonly int
LatentEncodingResult
readonly struct
Latent space encoding result.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
CompressionRatio
readonly float
LatentChannels
readonly int
LatentHeight
readonly int
LatentRepresentation
readonly float[]
LatentWidth
readonly int
ReconstructionError
readonly float
MultiAngleImageData
readonly struct
Multi-angle image data for comprehensive 3D reconstruction.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
CameraParams
readonly CameraParameters
CameraPosition
readonly GTVector3
CameraRotation
readonly GTVector3
DepthData
readonly float[]
ImageData
readonly float[]
ImageHeight
readonly int
ImageWidth
readonly int
Timestamp
readonly float
NoisePredictionResult
readonly struct
Noise prediction result from U-Net.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
Confidence
readonly float
HiddenStates
readonly float[]
PredictedNoise
readonly float[]
Timestep
readonly int
ONNXModelData
readonly struct
ONNX model data.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
HasMetadata
readonly bool
ModelData
readonly byte[]
ModelType
readonly ModelType
Precision
readonly Precision
PhysicsSDF
readonly struct
Physics-calculable SDF representation.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
GridResolution
readonly int
GridSpacing
readonly GTVector3
MaxBounds
readonly GTVector3
MinBounds
readonly GTVector3
SDFGrid
readonly float[]
SurfaceGrid
readonly bool[]
PointCloud3D
readonly struct
Point cloud 3D representation.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
Colors
readonly GTVector3[]
Densities
readonly float[]
Normals
readonly GTVector3[]
Points
readonly GTVector3[]
ReconstructionQuality
readonly struct
Reconstruction quality metrics.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
Accuracy
readonly float
Completeness
readonly float
Coverage
readonly float
TemporalStability
readonly float
TextEmbeddingResult
readonly struct
Text embedding result for conditioning.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
AttentionMask
readonly float[]
EmbeddingDim
readonly int
Embeddings
readonly float[]
PositionalEncoding
readonly float[]
SequenceLength
readonly int
TrainingMetrics
readonly struct
Training metrics for monitoring progress.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
Accuracy
readonly float
Epoch
readonly int
Loss
readonly float
PSNR
readonly float
SSIM
readonly float
VideoFrameData
readonly struct
Video frame data for temporal analysis.
Source: IntelligenceAlgorithmsDiffusion.cs
Constants and Fields
CameraPosition
readonly GTVector3
CameraRotation
readonly GTVector3
DepthData
readonly float[]
FrameNumber
readonly int
ImageData
readonly float[]
ImageHeight
readonly int
ImageWidth
readonly int
Timestamp
readonly float
GTOS.IntelligenceAlgorithms.Geospatial
AreaResult
readonly struct
Polygon area calculation result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Area
readonly double
Centroid
readonly GeoCoordinate
Perimeter
readonly double
Unit
readonly SpatialUnit
VertexCount
readonly int
BearingResult
readonly struct
Bearing and azimuth result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Distance
readonly double
FinalBearing
readonly double
InitialBearing
readonly double
Unit
readonly SpatialUnit
BufferResult
readonly struct
Buffer zone result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Area
readonly double
BufferDistance
readonly double
BufferPolygon
readonly GeoCoordinate[]
SegmentCount
readonly int
ClusterResult
readonly struct
Spatial cluster result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Algorithm
readonly ClusteringAlgorithm
ClusterCenter
readonly GeoCoordinate
ClusterId
readonly int
ClusterRadius
readonly double
PointCount
readonly int
ContainmentResult
readonly struct
Point-in-polygon test result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Confidence
readonly float
DistanceToEdge
readonly double
IsContained
readonly bool
Relation
readonly SpatialRelation
DistanceResult
readonly struct
Distance calculation result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Accuracy
readonly float
Bearing
readonly double
Distance
readonly double
Method
readonly DistanceMethod
Unit
readonly SpatialUnit
GeoCoordinate
readonly struct
Geographic coordinate (latitude, longitude, altitude).
WHAT IT CONTAINS:
- Latitude in degrees (-90 to +90)
- Longitude in degrees (-180 to +180)
- Altitude in meters (optional)
USE THIS FOR:
- GPS locations
- Site coordinates
- Asset tracking
- Mapping
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
Altitude
readonly double
Latitude
readonly double
Longitude
readonly double
GeospatialML
static class
Geospatial intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsGeospatial.cs
IntersectionResult
readonly struct
Intersection analysis result.
Source: IntelligenceAlgorithmsGeospatial.cs
Constants and Fields
IntersectionArea
readonly double
IntersectionPoints
readonly GeoCoordinate[]
Intersects
readonly bool
Relation
readonly SpatialRelation
GTOS.IntelligenceAlgorithms.Graph
CentralityResult
readonly struct
Centrality measure result.
WHAT IT CONTAINS:
- Centrality score for a node
- Rank among all nodes
- Normalized score
USE THIS FOR:
- Identifying critical nodes
- Ranking importance
- Vulnerability assessment
- Network resilience
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
CentralityScore
readonly double
Confidence
readonly float
NormalizedScore
readonly double
Rank
readonly int
Type
readonly CentralityType
ClusteringResult
readonly struct
Clustering coefficient result.
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
GlobalClustering
readonly double
LocalClustering
readonly double
PossibleTriangles
readonly int
TriangleCount
readonly int
CommunityResult
readonly struct
Community detection result.
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
CommunityId
readonly int
CommunitySize
readonly int
InternalDensity
readonly double
Modularity
readonly double
Quality
readonly CommunityQuality
ConnectivityResult
readonly struct
Connectivity analysis result.
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
AveragePathLength
readonly double
ComponentCount
readonly int
Diameter
readonly int
LargestComponentSize
readonly int
Level
readonly ConnectivityLevel
FlowResult
readonly struct
Network flow result.
WHAT IT CONTAINS:
- Maximum flow value
- Flow distribution across edges
- Bottleneck identification
USE THIS FOR:
- Capacity planning
- Bottleneck analysis
- Resource allocation
- Supply chain optimization
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
BottleneckCapacity
readonly double
BottleneckEdge
readonly int
MaxFlowValue
readonly double
Type
readonly FlowType
UtilizedCapacity
readonly double
GraphML
static class
Graph and network intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsGraph.cs
LinkPredictionResult
readonly struct
Link prediction result.
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
Confidence
readonly float
Node1
readonly int
Node2
readonly int
PredictionScore
readonly double
RecommendConnection
readonly bool
PathResult
readonly struct
Shortest path result.
WHAT IT CONTAINS:
- Path nodes and total distance
- Path cost and validity
USE THIS FOR:
- Route planning
- Network navigation
- Dependency ordering
- Flow optimization
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
PathExists
readonly bool
PathLength
readonly int
PathNodes
readonly int[]
TotalCost
readonly double
TotalDistance
readonly double
SpanningTreeResult
readonly struct
Spanning tree result.
Source: IntelligenceAlgorithmsGraph.cs
Constants and Fields
EdgeCount
readonly int
IsMinimal
readonly bool
TotalWeight
readonly double
TreeEdges
readonly int[]
GTOS.IntelligenceAlgorithms.Kinematics
CollisionAnalysis
readonly struct
Collision analysis result.
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
Confidence
readonly float
ImpactEnergy
readonly double
ImpactPoint
readonly Vector3D
ImpactVelocity
readonly Vector3D
Risk
readonly CollisionRisk
TimeToImpact
readonly double
KinematicsML
static class
Kinematics and motion intelligence atomic calculations.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsKinematics.cs
KinematicState
readonly struct
Velocity and acceleration estimation result.
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
Acceleration
readonly Vector3D
Confidence
readonly float
Position
readonly Vector3D
Timestamp
readonly double
Velocity
readonly Vector3D
MotionAnomalyResult
readonly struct
Motion anomaly detection result.
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
AnomalyScore
readonly double
AnomalySeverity
readonly float
AnomalyType
readonly MotionAnomalyType
AnomalyVector
readonly Vector3D
DetectionTimestamp
readonly double
MotionClassification
readonly struct
Motion pattern classification result.
WHAT IT CONTAINS:
- What type of motion (Linear, Circular, etc.)
- How confident (0-100%)
- Quality assessment (Nominal, Warning, Critical)
- Key motion parameters
USE THIS FOR:
- Equipment behavior verification
- Motion quality control
- Fault diagnosis
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
Amplitude
readonly double
Frequency
readonly double
PrimaryConfidence
readonly float
PrimaryType
readonly MotionType
Quality
readonly MotionQuality
SecondaryConfidence
readonly float
SecondaryType
readonly MotionType
TrajectoryPrediction
readonly struct
Trajectory prediction result with position, velocity, and confidence.
WHAT IT CONTAINS:
- Where the object will be (predicted position)
- How fast it will be moving (predicted velocity)
- How confident we are (0-100%)
- When collision will occur (-1 if none)
USE THIS FOR:
- Robot path planning
- Collision avoidance
- Timing coordination
- Safety zone violations
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
ConfidencePercent
readonly float
PredictedPosition
readonly Vector3D
PredictedVelocity
readonly Vector3D
PredictionHorizonSeconds
readonly double
TimeToCollision
readonly double
Vector3D
readonly struct
3D position vector for kinematics calculations.
MIL-SPEC: Zero-allocation value type.
Source: IntelligenceAlgorithmsKinematics.cs
Constants and Fields
X
readonly double
Y
readonly double
Z
readonly double
GTOS.IntelligenceAlgorithms.Materials
CompositionResult
readonly struct
Chemical composition analysis result.
Source: IntelligenceAlgorithmsMaterials.cs
Constants and Fields
Concentrations
readonly double[]
MoleFractions
readonly double[]
Purity
readonly double
TotalMass
readonly double
DefectResult
readonly struct
Material defect analysis result.
Source: IntelligenceAlgorithmsMaterials.cs
Constants and Fields
DefectCount
readonly int
DefectDensity
readonly double
DefectSizes
readonly double[]
Quality
readonly QualityClass
KineticsResult
readonly struct
Reaction kinetics result.
Source: IntelligenceAlgorithmsMaterials.cs
Constants and Fields
ActivationEnergy
readonly double
HalfLife
readonly double
RateConstant
readonly double
ReactionOrder
readonly double
Type
readonly ReactionType
MaterialPropertyResult
readonly struct
Material property prediction result.
Source: IntelligenceAlgorithmsMaterials.cs
Constants and Fields
Conductivity
readonly double
Confidence
readonly double
Density
readonly double
Elasticity
readonly double
Strength
readonly double
Type
readonly MaterialType
MaterialsML
static class
Materials and chemistry intelligence atomic calculations.
Source: IntelligenceAlgorithmsMaterials.cs
GTOS.IntelligenceAlgorithms.Medical
ECGResult
readonly struct
ECG analysis result.
Source: IntelligenceAlgorithmsMedical.cs
Constants and Fields
HeartRate
readonly double
QTInterval
readonly double
Quality
readonly SignalQuality
Rhythm
readonly ECGRhythm
RRInterval
readonly double
MedicalML
static class
Medical and biosignal intelligence atomic calculations.
Source: IntelligenceAlgorithmsMedical.cs
RiskAssessmentResult
readonly struct
Patient risk assessment result.
Source: IntelligenceAlgorithmsMedical.cs
Constants and Fields
Confidence
readonly double
ContributingFactors
readonly double[]
Level
readonly RiskLevel
Score
readonly double
SleepStageResult
readonly struct
Sleep stage classification result.
Source: IntelligenceAlgorithmsMedical.cs
Constants and Fields
Confidence
readonly double
Duration
readonly double
SleepEfficiency
readonly double
Stage
readonly int
VitalSignsResult
readonly struct
Vital signs analysis result.
Source: IntelligenceAlgorithmsMedical.cs
Constants and Fields
DiastolicBP
readonly double
HeartRate
readonly double
Risk
readonly RiskLevel
SpO2
readonly double
SystolicBP
readonly double
Temperature
readonly double
GTOS.IntelligenceAlgorithms.NLP
EntityMatch
readonly struct
Named entity extraction result.
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
Confidence
readonly float
Length
readonly int
StartIndex
readonly int
Text
readonly string
Type
readonly EntityType
KeywordResult
readonly struct
Keyword extraction result.
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
FirstPosition
readonly int
Frequency
readonly int
ImportanceScore
readonly double
Keyword
readonly string
LanguageDetectionResult
readonly struct
Language detection result.
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
CharacterCount
readonly int
Confidence
readonly float
DetectedLanguage
readonly LanguageCode
NGramResult
readonly struct
N-gram analysis result.
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
Frequency
readonly int
NGram
readonly string
NGramSize
readonly int
NormalizedFrequency
readonly double
NLPML
static class
Natural language processing intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsNLP.cs
SentimentResult
readonly struct
Sentiment analysis result.
WHAT IT CONTAINS:
- Polarity (positive/negative/neutral)
- Confidence score
- Emotional intensity
USE THIS FOR:
- Analyzing feedback
- Prioritizing negative reports
- Tracking sentiment trends
- Quality assessment
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
Confidence
readonly float
Intensity
readonly float
NegativeScore
readonly float
NeutralScore
readonly float
Polarity
readonly SentimentPolarity
PositiveScore
readonly float
TextSimilarityResult
readonly struct
Text similarity comparison result.
WHAT IT CONTAINS:
- Similarity score (0-1)
- Metric used
- Common tokens count
USE THIS FOR:
- Finding duplicate reports
- Matching descriptions
- Clustering similar documents
- Search relevance
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
CommonTokens
readonly int
Confidence
readonly float
Metric
readonly SimilarityMetric
SimilarityScore
readonly float
TotalTokens
readonly int
TfIdfResult
readonly struct
TF-IDF term importance result.
Source: IntelligenceAlgorithmsNLP.cs
Constants and Fields
DocumentFrequency
readonly int
IdfScore
readonly double
Term
readonly string
TermFrequency
readonly int
TfIdfScore
readonly double
TfScore
readonly double
GTOS.IntelligenceAlgorithms.ProcessControl
PerformanceResult
readonly struct
Control performance result.
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
IAE
readonly double
ISE
readonly double
Overshoot
readonly double
RiseTime
readonly double
SettlingTime
readonly double
SteadyStateError
readonly double
PIDResult
readonly struct
PID controller output result.
WHAT IT CONTAINS:
- Control output signal
- P, I, D components
- Error terms
USE THIS FOR:
- Temperature control
- Speed regulation
- Position control
- Process automation
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
ControlOutput
readonly double
DerivativeTerm
readonly double
Error
readonly double
IntegralTerm
readonly double
OutputSaturated
readonly bool
ProportionalTerm
readonly double
ProcessControlML
static class
Process control intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsProcessControl.cs
StabilityResult
readonly struct
Stability analysis result.
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
Eigenvalues
readonly double[]
GainMargin
readonly double
PhaseMargin
readonly double
StabilityIndex
readonly double
Status
readonly StabilityStatus
StateSpaceResult
readonly struct
State-space control result.
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
ControlInput
readonly double[]
Cost
readonly double
Output
readonly double[]
Stability
readonly StabilityStatus
State
readonly double[]
SystemIDResult
readonly struct
System identification result.
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
DeadTime
readonly double
FitQuality
readonly double
Gain
readonly double
Parameters
readonly double[]
TimeConstant
readonly double
Type
readonly ProcessType
TuningResult
readonly struct
PID tuning result.
Source: IntelligenceAlgorithmsProcessControl.cs
Constants and Fields
ExpectedPerformance
readonly double
Kd
readonly double
Ki
readonly double
Kp
readonly double
Method
readonly TuningMethod
GTOS.IntelligenceAlgorithms.Recommendation
ColdStartResult
readonly struct
Cold start recommendation result.
Source: IntelligenceAlgorithmsRecommendation.cs
Constants and Fields
Coverage
readonly double
ItemIds
readonly int[]
Scores
readonly double[]
RankingResult
readonly struct
Item ranking result.
Source: IntelligenceAlgorithmsRecommendation.cs
Constants and Fields
Diversity
readonly double
ItemIds
readonly int[]
Novelty
readonly double
Scores
readonly double[]
RecommendationML
static class
Recommendation intelligence atomic calculations.
Source: IntelligenceAlgorithmsRecommendation.cs
RecommendationResult
readonly struct
Recommendation result for single item.
Source: IntelligenceAlgorithmsRecommendation.cs
Constants and Fields
Confidence
readonly double
ItemId
readonly int
Reasons
readonly string[]
Score
readonly double
Type
readonly RecommendationType
SimilarityResult
readonly struct
User similarity result.
Source: IntelligenceAlgorithmsRecommendation.cs
Constants and Fields
CommonItems
readonly int
Similarity
readonly double
UserId
readonly int
GTOS.IntelligenceAlgorithms.ReinforcementLearning
BanditResult
readonly struct
Multi-armed bandit result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
ArmCounts
readonly int[]
ArmValues
readonly double[]
ExpectedReward
readonly double
RegretBound
readonly double
SelectedArm
readonly int
ExplorationResult
readonly struct
Exploration result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
IsExploration
readonly bool
SelectedAction
readonly int
SelectionProbability
readonly double
Strategy
readonly ExplorationStrategy
MonteCarloResult
readonly struct
Monte Carlo return result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
DiscountedReturn
readonly double
EpisodeLength
readonly int
EpisodeRewards
readonly double[]
Return
readonly double
PolicyResult
readonly struct
Policy evaluation result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
AverageReturn
readonly double
OptimalPolicy
readonly int[]
StateValues
readonly double[]
Status
readonly ConvergenceStatus
Type
readonly PolicyType
QLearningResult
readonly struct
Q-learning result.
WHAT IT CONTAINS:
- Q-values for state-action pairs
- Optimal action
- Value estimate
- Learning progress
USE THIS FOR:
- Process optimization decisions
- Resource allocation
- Sequential planning
- Adaptive control
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
Episodes
readonly int
LearningProgress
readonly double
MaxQValue
readonly double
OptimalAction
readonly int
QValues
readonly double[]
ReinforcementLearningML
static class
Reinforcement learning intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
RewardShapingResult
readonly struct
Reward shaping result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
OriginalReward
readonly double
PotentialDifference
readonly double
ShapedReward
readonly double
Type
readonly RewardType
TDResult
readonly struct
Temporal difference learning result.
Source: IntelligenceAlgorithmsReinforcementLearning.cs
Constants and Fields
Algorithm
readonly LearningAlgorithm
LearningRate
readonly double
TDError
readonly double
UpdatedValue
readonly double
GTOS.IntelligenceAlgorithms.SignalProcessing
Complex
readonly struct
Complex number for frequency domain.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
Imaginary
readonly double
Real
readonly double
CorrelationResult
readonly struct
Correlation analysis result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
Correlation
readonly double[]
LagAtMaxCorrelation
readonly int
MaxCorrelation
readonly double
MaxLag
readonly int
EnvelopeResult
readonly struct
Signal envelope result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
AverageAmplitude
readonly double
EnvelopeModulation
readonly double
LowerEnvelope
readonly double[]
MaxAmplitude
readonly double
UpperEnvelope
readonly double[]
FFTResult
readonly struct
FFT analysis result.
WHAT IT CONTAINS:
- Frequency spectrum (magnitude and phase)
- Dominant frequencies
- Spectral power
USE THIS FOR:
- Vibration analysis
- Audio spectrum
- Frequency identification
- Harmonic analysis
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
DominantFrequency
readonly double
Magnitudes
readonly double[]
SampleCount
readonly int
Spectrum
readonly Complex[]
TotalPower
readonly double
FilterResult
readonly struct
Digital filter result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
AttenuationDB
readonly double
CutoffFrequency
readonly double
FilteredSignal
readonly double[]
Quality
readonly SignalQuality
Type
readonly FilterType
PeakResult
readonly struct
Peak detection result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
AveragePeakHeight
readonly double
AveragePeakSpacing
readonly double
PeakCount
readonly int
PeakIndices
readonly int[]
PeakValues
readonly double[]
PowerSpectrumResult
readonly struct
Spectral power density result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
BandwidthHz
readonly double
Frequencies
readonly double[]
PeakFrequency
readonly double
PeakPower
readonly double
PowerDensity
readonly double[]
SignalProcessingML
static class
Signal processing intelligence atomic calculations.
All methods are MIL-SPEC compliant with minimal allocation.
Source: IntelligenceAlgorithmsSignalProcessing.cs
SignalStatistics
readonly struct
Signal statistics result.
Source: IntelligenceAlgorithmsSignalProcessing.cs
Constants and Fields
CrestFactor
readonly double
Kurtosis
readonly double
Mean
readonly double
PeakToPeak
readonly double
Quality
readonly SignalQuality
RMS
readonly double
StandardDeviation
readonly double
GTOS.IntelligenceAlgorithms.TimeSeries
AutoCorrelationResult
readonly struct
Auto-correlation analysis result.
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
Correlations
readonly double[]
DetectedPeriod
readonly SeasonalityType
DominantLag
readonly int
SignificantLags
readonly int[]
Strength
readonly float
ChangePointResult
readonly struct
Change point detection result.
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
ChangePointIndex
readonly int
ChangePointTime
readonly double
Confidence
readonly float
MagnitudeChange
readonly double
Type
readonly ChangePointType
ValueAfter
readonly double
ValueBefore
readonly double
ForecastResult
readonly struct
Forecast result with prediction and confidence intervals.
WHAT IT CONTAINS:
- Point forecast: Best estimate
- Confidence intervals: Upper/lower bounds
- Forecast horizon and confidence level
- Trend and seasonal components
USE THIS FOR:
- Future value prediction
- Uncertainty quantification
- Risk assessment
- Planning with confidence bounds
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
Confidence
readonly ConfidenceLevel
HorizonSteps
readonly int
LowerBound
readonly double
PointForecast
readonly double
SeasonalComponent
readonly double
Timestamp
readonly double
TrendComponent
readonly double
UpperBound
readonly double
LeadLagResult
readonly struct
Lead-lag relationship result.
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
Confidence
readonly float
Correlation
readonly double
IsLeading
readonly bool
LeadLagOffset
readonly int
PredictiveStrength
readonly double
SeasonalDecomposition
readonly struct
Seasonal decomposition result (Trend + Seasonal + Residual).
WHAT IT CONTAINS:
- Trend component: Long-term direction
- Seasonal component: Repeating patterns
- Residual component: Random noise
- Seasonality period and strength
USE THIS FOR:
- Understanding data components
- Removing seasonality for analysis
- Forecasting with seasonal patterns
- Data quality assessment
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
DetectedSeasonality
readonly SeasonalityType
ResidualValue
readonly double
SeasonalPeriod
readonly int
SeasonalStrength
readonly float
SeasonalValue
readonly double
TrendStrength
readonly float
TrendValue
readonly double
StreamAnomalyResult
readonly struct
Stream anomaly detection result.
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
ActualValue
readonly double
AnomalyScore
readonly double
Deviation
readonly double
ExpectedValue
readonly double
IsAnomaly
readonly bool
Timestamp
readonly double
Type
readonly StreamAnomalyType
TimeSeriesML
static class
Time series and forecasting intelligence atomic calculations.
All methods are zero-allocation and MIL-SPEC compliant.
Source: IntelligenceAlgorithmsTimeSeries.cs
TrendAnalysis
readonly struct
Trend analysis result.
Source: IntelligenceAlgorithmsTimeSeries.cs
Constants and Fields
Confidence
readonly float
Intercept
readonly double
RSquared
readonly double
Slope
readonly double
Strength
readonly float
Type
readonly TrendType
Generated from GTOS Savants source -- 2026-03-22
