@codai/cbd/dist/features/ml-integration.d.ts

Codai Better Database - High-Performance Vector Memory System with HPKV-inspired architecture and MCP server

/**
 * Machine Learning Integration
 * Custom embedding models, model inference, auto-ML, and predictive analytics
 */
import { EventEmitter } from 'events';
interface MLModelConfig {
    modelType: 'embedding' | 'classification' | 'regression' | 'clustering' | 'anomaly-detection';
    modelPath?: string;
    apiEndpoint?: string;
    apiKey?: string;
    maxTokens?: number;
    temperature?: number;
}
interface ModelInferencePipeline {
    id: string;
    name: string;
    models: MLModelConfig[];
    preprocessing?: string[];
    postprocessing?: string[];
    caching: boolean;
}
interface PredictiveAnalytics {
    patternRecognition: boolean;
    anomalyDetection: boolean;
    trendForecasting: boolean;
    userBehaviorPrediction: boolean;
    performanceOptimization: boolean;
}
interface AutoMLConfig {
    enabled: boolean;
    featureEngineering: boolean;
    hyperparameterTuning: boolean;
    modelSelection: boolean;
    ensembleMethods: boolean;
}
declare class MachineLearningIntegration extends EventEmitter {
    private openai;
    private modelRegistry;
    private inferencePipelines;
    private modelCache;
    private predictiveEngine;
    private autoMLEngine;
    private performanceTracker;
    constructor(config: {
        openaiApiKey: string;
        modelConfigs?: MLModelConfig[];
        predictiveAnalytics?: PredictiveAnalytics;
        autoML?: AutoMLConfig;
        cacheSize?: number;
    });
    private initializeMLIntegration;
    /**
     * Custom Embedding Model Support
     */
    generateCustomEmbedding(text: string, modelId?: string, options?: {
        dimensions?: number;
        normalize?: boolean;
        batchSize?: number;
    }): Promise<{
        embedding: number[];
        modelUsed: string;
        processingTime: number;
        confidence: number;
    }>;
    /**
     * Model Inference Pipeline
     */
    runInferencePipeline(pipelineId: string, input: any, options?: {
        caching?: boolean;
        timeout?: number;
        retries?: number;
    }): Promise<{
        result: any;
        pipeline: string;
        executionTime: number;
        modelsUsed: string[];
        confidence: number;
    }>;
    /**
     * Auto-ML Feature Engineering
     */
    performAutoML(dataset: any[], target: string, options?: {
        taskType?: 'classification' | 'regression' | 'clustering';
        maxTime?: number;
        modelTypes?: string[];
        crossValidation?: number;
    }): Promise<{
        bestModel: any;
        performance: any;
        features: string[];
        hyperparameters: any;
        executionTime: number;
    }>;
    /**
     * Predictive Analytics
     */
    performPredictiveAnalysis(data: any[], analysisType: 'pattern-recognition' | 'anomaly-detection' | 'trend-forecasting' | 'user-behavior', options?: any): Promise<{
        predictions: any[];
        confidence: number;
        insights: string[];
        recommendations: string[];
        executionTime: number;
    }>;
    /**
     * Model Management
     */
    registerModel(modelId: string, config: MLModelConfig): void;
    createInferencePipeline(pipelineId: string, pipeline: ModelInferencePipeline): void;
    getModelPerformance(modelId: string): Promise<any>;
    private generateCustomAPIEmbedding;
    private normalizeVector;
    private applyPreprocessing;
    private applyPostprocessing;
    private applyPreprocessingStep;
    private applyPostprocessingStep;
    private runSingleModel;
    private setupPerformanceMonitoring;
}
export { MachineLearningIntegration, MLModelConfig, ModelInferencePipeline, PredictiveAnalytics, AutoMLConfig };
//# sourceMappingURL=ml-integration.d.ts.map