RankedBatchStrategy<T, TInput, TOutput>
Models & Types · AiDotNet.ActiveLearning.Batch
Simple ranked batch selection strategy with diversity filtering.
Implements: IBatchStrategy<T, TInput, TOutput>
For Beginners
This is the simplest batch selection strategy. It ranks samples by their informativeness scores and selects the top-k, optionally filtering out samples that are too similar to already-selected ones.
How It Works
How It Works:
Trade-offs:
Constructors
| Constructor | Summary |
|---|---|
RankedBatchStrategy() |
Initializes a new RankedBatchStrategy with default settings. |
RankedBatchStrategy(double minDiversityThreshold, double diversityTradeoff) |
Initializes a new RankedBatchStrategy with specified parameters. |
Properties
| Property | Summary |
|---|---|
T DiversityTradeoff { get; set; } |
|
string Name { get; } |
Methods
| Method | Summary |
|---|---|
T ComputeDiversity(TInput sample1, TInput sample2) |
|
int[] SelectBatch(int[] candidateIndices, Vector<T> scores, IDataset<T, TInput, TOutput> unlabeledPool, int batchSize) |