# What is Accuracy? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Mathematical Foundations  
> **Canonical Citation:** [https://spidits.com/ai-glossary/accuracy](https://spidits.com/ai-glossary/accuracy)

## Definition
Accuracy is a classification metric measuring the fraction of total predictions that the model got correct, calculated as the sum of correct predictions divided by all predictions.

## Architectural Mechanics & Deep Dive
In machine learning, accuracy is a straightforward performance metric calculated as the ratio of correctly predicted instances to total instances. While intuitive, it is highly deceptive when applied to imbalanced datasets. For example, in fraud detection where only 1% of transactions are fraudulent, a model that classifies everything as non-fraudulent achieves 99% accuracy while failing completely at its task. To build robust models, practitioners supplement accuracy with confusion matrix metrics such as precision, recall, and the F1-score, which evaluate performance per class.

## Industry Applications & Real-World Use Cases
General model evaluation, baseline performance tracking, and accuracy benchmarks.

## Frequently Asked Questions

### What is the formula for classification accuracy?
`Accuracy = (True Positives + True Negatives) / Total Predictions`.

### Why can accuracy be a misleading metric?
On highly imbalanced datasets, accuracy fails. If a dataset has 99% negative cases and 1% positive cases, a model that classifies everything as negative is 99% accurate but is completely useless at detecting positive events.

## Related Terminology
- [precision](https://spidits.com/ai-glossary/precision)
- [recall](https://spidits.com/ai-glossary/recall)
- [f1-score](https://spidits.com/ai-glossary/f1-score)
- [imbalanced-datasets](https://spidits.com/ai-glossary/imbalanced-datasets)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*