Artificial Intelligence Fundamentals and Project Cycle

A comprehensive set of practice flashcards covering basic AI terminology, the domains of AI, the detailed AI project cycle, and ethical considerations surrounding artificial intelligence.

Artificial Intelligence (AI)

The ability of machines to perform tasks that typically require human-like intelligence, such as decision-making, visual perception, and speech recognition.

John McCarthy

The researcher who coined the term 'Artificial Intelligence' in his proposal for the 1956 Dartmouth Conference.

Alan Turing

A pioneer who proposed a test in 1950 to check a machine's ability to exhibit intelligent behavior similar to human intelligence.

Data Domain

The AI domain focusing on the fuel of AI; it relates to the storage and processing of massive sets of information known as Big Data.

Computer Vision (CV)

A domain of AI that teaches machines to collect information from digital images and videos to make sense of them similarly to human vision.

Natural Language Processing (NLP)

The ability of computers to understand, interpret, and process human text and verbal language.

AI Project Cycle

A structured framework consisting of five stages: problem scoping, data acquisition, data exploration, modelling, and evaluation.

Problem Scoping

The first stage of the AI project cycle where clear goals are set and the objectives of the project are outlined.

4Ws Problem Canvas

A structured framework for problem scoping that identifies the Who (stakeholders), What (nature of the problem), Where (location/context), and Why (value of the solution).

Stakeholders

People who are directly or indirectly affected by a problem and who will benefit from the implemented AI solution.

Data Acquisition

The second stage of the project cycle focused on collecting relevant, accurate, and reliable data from authentic sources.

Training Data

The initial dataset used to train an AI model to identify patterns and perform specific tasks.

Testing Data

A separate dataset that the algorithm hasn't seen before, used to evaluate the model's performance in real-world situations.

Data Features

Descriptions of the type of information that will be collected in response to the problem statement.

Data Exploration

The process of analyzing collected data using visualization methods like charts and graphs to interpret patterns and trends.

System Maps

Visual diagrams that help understand the different parts or elements of an AI project and how they are interconnected.

Modelling

The stage where an appropriate AI model or algorithm is selected and trained with acquired data to generate predictions.

Rule-based Approach

An AI modelling approach where the relationships or patterns in data are predefined by the developer using specific rules.

Learning-based Approach

Also known as Machine Learning, this approach allows the machine to learn on its own from data without being explicitly programmed.

Decision Tree

A graphical representation used in rule-based approaches for making decisions based on branching conditions leading to leaf nodes (outcomes).

Supervised Learning

A learning model where the algorithm is provided with labeled datasets to learn a mapping between input data and output labels.

Classification

A type of supervised learning used for discrete or categorical datasets to group data into specific labeled categories.

Regression

A supervised learning model used to predict continuous or numerical values within a certain range, such as future stock prices.

Unsupervised Learning

An AI model that works on unlabeled datasets to discover patterns, trends, or clusters on its own.

Clustering

An unsupervised approach where the machine groups data into different clusters based on self-generated algorithms for similar patterns.

Dimensionality Reduction

A technique used to simplify complex, high-dimensional data by reducing its dimensions, even at the cost of losing some information.

Reinforcement Learning (RL)

A trial-and-error learning method where an agent interacts with an environment and receives rewards for good actions and penalties for bad ones.

Evaluation

The process of assessing model reliability and efficiency by comparing generated output with expected outcomes using testing data.

True Positive (TP)

An evaluation parameter where the model's positive prediction matches the positive reality (there is no error).

False Positive (FP)

Known as a Type 1 error, this occurs when the model predicts 'YES' but the actual reality is 'NO'.

False Negative (FN)

Known as a Type 2 error, this occurs when the model predicts 'NO' but the actual reality is 'YES'.

Deployment

The final step in the AI project cycle where the trained and validated model is implemented in a real-world scenario.

AI Ethics

Principles and moral guidelines governing the development and deployment of AI to ensure responsible use of the technology.

AI Bias

Unfair favoritism or prejudice towards certain individuals or groups caused by algorithms trained on biased or unrepresentative data.

Opacity / Black Box Effect

A critical AI issue where the inner workings or decision-making process of a system is not transparent to the user.

Singularity

The hypothesis that if AI reaches human-level intelligence, it may one day overtake or completely control the human species.