liveBook · Manning

Mathematical representation

As explained in The Art of AI Product Development

A numerical modality is data represented as numeric values that computational models can accept as input.

A numerical modality represents data as numeric values that computational models can accept as input. Although raw data may arrive as text, images, video, or audio, an AI model ultimately requires a numeric form. section 3.1.1

This transformation matters when mapping an AI solution: different raw modalities call for different techniques, while multimodal systems combine several modalities before or during model processing. section 3.1.1

Common relationships include:

  • Textual data can support language tasks such as sentiment analysis, translation, summarization, and question answering. section 3.1.1
  • Visual data can support image classification, object detection, and image generation through computer vision. section 3.1.1
  • One-hot encoding provides a simple algebraic representation of words as numeric data. section 3.1.1
Figure 3.4 One-hot encoding provides a simple numerical (algebraic) representation of words.
A screenshot of a computer

AI-generated content may be incorrect.

Use the figure to see how one-hot encoding turns words into a simple numerical representation. section 3.1.1

How The Art of AI Product Development covers it

How deep it goes
Defines the numeric-input requirement and transformation need, but stops short of implementation detail.
Examples
Uses text, images, video, audio, and one-hot encoding as examples of data or numerical representation.
What you can do after
Evaluate how raw data must be transformed before an AI model can process it.
How it connects
Connects numerical representation to raw data modalities, natural language processing, computer vision, and multimodal AI.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of The Art of AI Product Development. The section links point to the parts of the book it draws on.