Table of Contents

Getting Started with AiDotNet

This guide will help you install AiDotNet and build your first AI model.

Prerequisites

Before installing AiDotNet, ensure you have:

  • .NET 10.0 SDK (or .NET 8.0+) (Download)
  • (Optional) NVIDIA GPU with CUDA for GPU acceleration

Installation

NuGet Package Manager

dotnet add package AiDotNet

Package Manager Console (Visual Studio)

Install-Package AiDotNet

PackageReference (csproj)

<PackageReference Include="AiDotNet" Version="*" />

Your First Model

Let's build a simple classification model:

using AiDotNet;
using AiDotNet.Classification;

// Sample data
var features = new double[][]
{
    new[] { 5.1, 3.5, 1.4, 0.2 },
    new[] { 7.0, 3.2, 4.7, 1.4 },
    new[] { 6.3, 3.3, 6.0, 2.5 }
};
var labels = new double[] { 0, 1, 2 };

// Build and train
var result = await new AiModelBuilder<double, double[], double>()
    .ConfigureModel(new RandomForestClassifier<double>(nEstimators: 100))
    .ConfigurePreprocessing()  // Auto-applies StandardScaler
    .BuildAsync(features, labels);

// Predict using the result object (facade pattern)
var prediction = result.Predict(new[] { 5.9, 3.0, 5.1, 1.8 });
Console.WriteLine($"Predicted class: {prediction}");

Core Concepts

AiModelBuilder

The AiModelBuilder is the main entry point for building models. It uses a fluent API:

var builder = new AiModelBuilder<T, TInput, TOutput>();

Where:

  • T - Numeric type (usually double or float)
  • TInput - Input data type (e.g., double[], Tensor<T>)
  • TOutput - Output type (e.g., double, int)

Configuration Methods

The builder provides many Configure* methods:

Method Purpose
ConfigureModel() Set the ML model/algorithm
ConfigureOptimizer() Set the optimizer (for neural networks)
ConfigurePreprocessing() Add data preprocessing steps
ConfigurePostprocessing() Add output postprocessing
ConfigureCrossValidation() Enable cross-validation
ConfigureGpuAcceleration() Enable GPU training

Building and Training

Call BuildAsync() with your training data:

var result = await builder.BuildAsync(features, labels);

The result contains:

  • Model - The trained model
  • CrossValidationResult - CV metrics (if enabled)
  • TrainingMetrics - Training statistics

Next Steps

GPU Acceleration

To enable GPU training:

builder.ConfigureGpuAcceleration(new GpuAccelerationConfig
{
    Enabled = true,
    DeviceId = 0  // Use first GPU
});

Requirements:

  • NVIDIA GPU with CUDA support
  • CUDA Toolkit 11.8+
  • cuDNN 8.6+

Need Help?

Edit this page