ContinualLearningOptions

Options & Configuration · AiDotNet.Models.Options

Represents configuration options for continual learning.

Inheritance: ModelOptions → ContinualLearningOptions

For Beginners

Continual learning (also called lifelong learning) allows models to learn from a sequence of tasks without forgetting what was learned before. This is important because standard neural networks suffer from “catastrophic forgetting” - when trained on new data, they tend to forget previously learned patterns.

How It Works

Typical Usage:

How to Choose a Strategy:

Constructors

Constructor Summary
ContinualLearningOptions()

Properties

Property Summary
ContinualLearningStrategyType Strategy { get; set; } Gets or sets the continual learning strategy to use.
ReplayBufferStrategy BufferStrategy { get; set; } Gets or sets the buffer management strategy for Experience Replay.
bool NormalizeImportance { get; set; } Gets or sets whether to normalize importance scores.
bool UseLateralConnections { get; set; } Gets or sets whether to use lateral connections in Progressive Neural Networks.
double Damping { get; set; } Gets or sets the damping constant for Synaptic Intelligence.
double DropoutRate { get; set; } Gets or sets the dropout rate for Monte Carlo Dropout in VCL.
double Gamma { get; set; } Gets or sets the decay factor for Online EWC.
double InitialLogVariance { get; set; } Gets or sets the initial log-variance for weight distributions in VCL.
double Lambda { get; set; } Gets or sets the regularization strength (lambda) for weight consolidation strategies.
double Margin { get; set; } Gets or sets the margin for gradient constraints in GEM.
double PruningRatio { get; set; } Gets or sets the pruning ratio for PackNet.
double ReplayRatio { get; set; } Gets or sets the ratio of replay samples to new samples in replay-based strategies.
double Temperature { get; set; } Gets or sets the temperature for knowledge distillation in LearningWithoutForgetting.
int FisherSampleSize { get; set; } Gets or sets the number of samples to use for Fisher Information estimation.
int MaxBufferSize { get; set; } Gets or sets the maximum buffer size for Experience Replay.
int MemorySize { get; set; } Gets or sets the maximum number of samples to store per task in memory-based strategies.
int NumMcSamples { get; set; } Gets or sets the number of Monte Carlo samples for VCL.
int ReplayBatchSize { get; set; } Gets or sets the batch size for generating pseudo-examples in Generative Replay.
int SampleSize { get; set; } Gets or sets the batch size to sample from memory for reference gradients in A-GEM.