Table 1.
Definitions of AI Terminology
| Term | Definition |
|---|---|
| Artificial intelligence | Branch of computer science dedicated to developing computer algorithms that emulate intelligent human behavior, such as learning, recognizing patterns, reasoning, solving problems, making decisions, and self-correcting. |
| Classification | A supervised learning method to predict class membership of an observation. |
| Deep learning | A subfield of machine learning that relies on neural networks with multiple layers to progressively extract higher-level features from raw data. |
| External validation | Validation of a model using data from a source that is different from the training data. |
| Ground truth | Correct labels (or true labels) for data, as determined by experts or other reference standards. |
| Hidden layer | A synthetic layer in a neural network between the input layer (ie, the features) and the output layer (ie, the prediction). |
| Internal validation | Validation of a model using data from the same source as the training data. |
| Machine learning | A subfield of artificial intelligence in which computers learn without being explicitly programmed. |
| Neural network | A multi-layer network that resembles the connectivity of neurons in the brain. |
| Overfitting | Occurs when a model is trained to predict the training dataset so well that it may fail to make a good prediction on new data. |
| Regression | A supervised learning method to predict output with continuous value. |
| Reinforcement learning | A type of machine learning in which the algorithm learns from positive and negative feedback without being taught. |
| Supervised learning | A type of machine learning in which the algorithm is provided with labeled training data. |
| Test set | A subset of the dataset that is used to evaluate the model. |
| Training set | A subset of the dataset that is used to develop the model. |
| Unsupervised learning | A type of machine learning in which the algorithm is provided with training data without corresponding labels. |
| Validation set | A subset of the dataset that is used to fine-tune the model’s parameters. |