✦ScienexySCIENCE KNOWLEDGE LIBRARY
Science reference · SCI-CS-003-061

Datasets and Data Structures in ML

A dataset is an organized collection of observations used to train, validate, and evaluate machine learning models. Its structure determines how information is represented and processed.

Explanation

A dataset is a collection of related data samples. Each sample represents an observation, such as a customer, image, transaction, or medical record. In a tabular dataset, rows usually represent samples and columns represent features or attributes. The target column contains the value a supervised learning model attempts to predict. Datasets may also contain numerical arrays, matrices, tensors, text sequences, images, audio recordings, or graph structures. The appropriate representation depends on the learning problem. For example, a spreadsheet is suitable for customer information, a tensor is commonly used to represent image batches, and a graph can represent relationships between users in a social network. Understanding dataset structure helps determine which preprocessing methods, algorithms, and evaluation strategies are appropriate.

Example or application

A house-price dataset may contain rows representing individual houses and columns representing area, number of bedrooms, location, and selling price. The first three columns are features, while selling price is the target.

Text

Machine learning systems learn patterns from structured or unstructured data organized into datasets.

Key Points

  • A dataset contains multiple related data samples.

  • Rows commonly represent observations in tabular data.

  • Columns commonly represent features and targets.

  • Datasets can be structured or unstructured.

  • Data representation influences model selection and preprocessing.

Applications

  • Customer behavior prediction

  • Image classification

  • Financial forecasting

  • Medical data analysis

  • Recommendation systems

Learning Outcomes

  • Define a dataset and identify its main components.

  • Distinguish between features and target variables.

  • Recognize common machine learning data structures.

  • Choose suitable data representations for different problems.

Practice Questions

  • What is a dataset in machine learning?

  • What is the difference between a feature and a target?

  • How does a matrix differ from a tensor?

  • Which data structure could represent relationships between social network users?

Related Concepts

  • Features

  • Target Variables

  • Structured Data

  • Unstructured Data

  • Tensors

Common Mistakes

  • Assuming every dataset must be a spreadsheet.

  • Confusing features with target variables.

  • Ignoring the relationship between data structure and model requirements.

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Géron, A. (2022). Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow.
  • Bishop, C. M. (2006). Pattern Recognition and Machine Learning.

Suggest a correction or source · More in Artificial Intelligence & Machine Learning