1/41
Looks like no tags are added yet.
Expert system
An intelligent computer program that uses knowledge and inference procedures to solve problems that are difficult enough to require significant human expertise for their solutions.
Knowledge base
A module containing a collection of factual and heuristic information in a given field.
Factual knowledge
Knowledge based on facts.
Heuristic knowledge
Knowledge based on practice, evaluation, experiences, and the ability to guess.
Inference engine
A component used to acquire insights from the information housed in the knowledge base by repeatedly applying rules to working memory until a goal state is produced or confirmed.
Control system
A functional element of an inference engine that determines the order of testing in the knowledge base rules.
Rule interpreter
A functional element of an inference engine that defines the Boolean application rules.
Explanation mechanism
A functional element of an inference engine that justifies the outcome to the user with the reasoning process.
Forward chaining
An inference strategy where the process moves from the facts of the case to a goal (conclusion) by matching each rule's condition (IF) with facts in working memory.
Backward chaining
An inference strategy where the engine attempts to match the assumed conclusion to the goal or sub-goal state with the conclusion (THEN) part of the rule.
User interface
The component of an expert system that helps users communicate with it and helps explain how it arrived at a particular recommendation.
Explanation facility
A component that allows the user to ask the expert system how it reached a particular conclusion and why a specific task (fact) is needed.
Knowledge acquisition
The process of eliciting, collecting, analyzing, modeling, and validating knowledge for knowledge engineering and management projects.
External interface
A module that allows an expert system to work with external files using programs written in conventional programming languages like C and C++.
Database
A collection of organized information used by the inference engine to hold data about current tasks while working on a problem.
Rule-based system architecture
The most common form of expert system architecture that uses knowledge in the form of production rules (if-then rules), also known as production systems.
Decision tree
A non-production system representation scheme structured in a top-to-bottom manner that takes input from an object's properties and outputs a Boolean decision.
Blackboard system architecture
A special type of knowledge-based system architecture that uses different knowledge sources communicating through a common information field.
Knowledge sources
Independent modules in a blackboard system that contain the knowledge needed for problem solving.
Blackboard
A global database in a blackboard system used for sharing input data, partial solutions, alternatives, and final solutions.
Control unit
A component in a blackboard system that determines which knowledge sources to execute for an optimal problem solution.
Neural networks
Computing systems modeled on the human brain's mesh-like network of interconnected processing elements called neurons.
Machine learning
The process of training a model to make useful predictions from data.
Machine learning model
A complex collection of numbers representing the mathematical relationship between elements of data that an ML system uses to make predictions.
Supervised learning
A machine learning category where the machine is provided with an established set of labeled data.
Regression
A supervised learning category where the model predicts the values of a continuous variable.
Classification
A supervised learning model that predicts the likelihood that an item belongs to a distinct set of categories.
Binary classification model
A classification model that outputs a value from a class containing only two values.
Multiclass classification model
A classification model that outputs a value from a class containing more than two values.
Unsupervised learning
A machine learning category where the model makes predictions based on unlabeled data that is not provided with a specific answer.
Clustering
A common unsupervised learning technique where items with common properties are grouped together.
Reinforcement learning
A machine learning category where the model learns through trial and error, receiving feedback to guide it to the best outcome.
Data
The driving force of machine learning, stored as related entries in datasets.
Datasets
Collections of individual examples or entries that contain features and a label.
Features
The input values in a dataset used to predict the label.
Label
The answer or output value that a supervised machine learning model has to predict.
Labeled examples
Dataset entries that contain both features and a label.
Unlabeled examples
Dataset entries that consist of features without labels.
Training
The process where a model discovers patterns and learns mathematical relationships by comparing predicted values to actual label values in a dataset.
Loss
The difference between a model's predicted value and the label's actual value during training.
Model evaluation
The process of testing a trained model using only the features of a labeled dataset and comparing its predictions to the true label values.
Inferences
Predictions made by a trained machine learning model on unlabeled examples.