Terms

Flashcards of key terms and definitions from the eHealth lecture.

Data

Raw facts or measurements without context (e.g., blood pressure 180/110).

Information

Data with meaning and context (e.g., blood pressure 180/110 is classified as "high").

Knowledge

Understanding what information means and what actions to take (e.g., high blood pressure needs treatment to reduce heart disease risk).

KDD (Knowledge Discovery in Databases)

The process of discovering useful knowledge from data through: data collection, preprocessing, pattern finding, and interpretation.

Narrative Data

Free-text clinical notes written by healthcare providers about patient symptoms, treatments, and plans.

Structured Text Data

Data that follows a predefined format, such as standardized forms or coded data.

Numerical Measurement Data

Quantitative values representing health parameters (e.g., blood pressure, temperature).

Signal Data

Continuous recordings of physiological processes over time (e.g., ECG, EEG).

Image Data

Visual representations of body structures (e.g., X-rays, CT scans, MRIs).

Nominal Scale

Categories with no inherent order (e.g., blood types A, B, AB, O).

Ordinal Scale

Categories with a meaningful order but unequal intervals (e.g., pain scale 1-10).

Interval Scale

Equal intervals between values but no true zero (e.g., temperature in Celsius).

Ratio Scale

Equal intervals with a meaningful zero point (e.g., weight, height).

PPG (Photoplethysmography)

Technology that measures blood volume changes in vessels using light to determine heart rate and other cardiovascular metrics.

ECG (Electrocardiogram)

Technology that measures the electrical activity of the heart using electrodes on the skin.

Green Light PPG

PPG technology using green light, which provides good signal-to-noise ratio but is affected by skin tone and limited in tissue penetration.

Infrared Light PPG

PPG technology using infrared light, which penetrates deeper into tissue and is less affected by skin tone.

HRV (Heart Rate Variability)

The variation in time intervals between successive heartbeats, reflecting autonomic nervous system function.

AVNN

Average of inter-beat intervals; a time-domain HRV metric.

SDNN

Standard deviation of inter-beat intervals; a time-domain HRV metric indicating overall HRV.

RMSSD

Root mean square of successive differences between normal heartbeats; a time-domain HRV metric reflecting parasympathetic activity.

PWV (Pulse Wave Velocity)

The speed at which the pressure wave moves along an artery, calculated as distance divided by time delay between pulse waves.

Data Mining

The process of discovering patterns in data that are non-trivial, hidden, and potentially useful.

Descriptive Data Mining

Finding patterns in existing data without making predictions (e.g., clustering, association rule mining).

Predictive Data Mining

Using historical data patterns to make predictions about future outcomes (e.g., classification, regression).

Supervised Learning

Machine learning approach that uses labeled training data to learn a model for prediction.

Unsupervised Learning

Machine learning approach that finds patterns in unlabeled data without predefined categories.

Clustering

Grouping similar data objects together while keeping dissimilar objects in different groups, without using predefined categories.

K-means Clustering

A partitioning method that divides data into k clusters by minimizing the distance between data points and their assigned cluster centers.

Partitioning Around Medoids (PAM)

A clustering algorithm similar to K-means but uses actual data points as cluster centers instead of calculated means.

Classification

Assigning items to predefined categories based on patterns learned from labeled training data.

Regression

Finding a formula to predict a numerical value based on other information.

Attribute Selection

The process of choosing only relevant features from data and ignoring irrelevant ones to improve model performance.

Attribute Construction

Creating new, more useful features from existing ones to help find patterns not visible in original data.

Confusion Matrix

A table showing prediction performance with True Positives, False Positives, True Negatives, and False Negatives.

Accuracy

The proportion of correct predictions: (TP + TN) / (TP + TN + FP + FN).

Sensitivity (True Positive Rate)

The proportion of actual positives correctly identified: TP / (TP + FN).

Specificity (True Negative Rate)

The proportion of actual negatives correctly identified: TN / (TN + FP).

Black-box Model

Model that provides predictions without explaining its reasoning process (e.g., neural networks, SVMs).

White-box Model

Model that provides interpretable decision processes (e.g., decision trees, rule-based systems).

Cross-Validation

A technique to evaluate model performance by partitioning data into multiple training and testing subsets.

K-fold Cross-Validation

Dividing data into k equal parts, using k-1 parts for training and 1 part for testing, repeating k times with a different test set each time.

Stratification

Ensuring each fold in cross-validation has approximately the same proportion of class values as the original dataset.

LOOCV (Leave-One-Out Cross-Validation)

A special case of k-fold cross-validation where k equals the number of examples, using all but one example for training.

Workflow

A representation of a sequence of operations, defined as a work process consisting of a set of tasks with specific ordering.

Task Set

The collection of all tasks in a workflow, denoted as T = {T₁, T₂, …, Tₘ}.

Precedence Matrix

A matrix P = (Pᵢⱼ)ₘₓₘ where Pᵢⱼ = 1 indicates task j must be completed before task i can start.

Conflict Matrix

A matrix C = (cᵢⱼ)ₘₓₘ where cᵢⱼ = 1 indicates tasks i and j cannot be performed simultaneously.

Precondition Set

For each task, the set of tasks that must be completed before it can start, denoted as A = (A(T₁), A(T₂), …, A(Tₘ)).

Task State 0

Task is not executable and not executed previously.

Task State 1

Task is executable and not executed previously.

Task State 2

Task is not executable and executed previously.

Task State 3

Task is executable and executed previously.

Initial State

The starting state of a workflow, denoted as S₀.

Bayes' Theorem

P(A|B) = [P(B|A) × P(A)] / [P(B|A) × P(A) + P(B|A^c) × P(A^c)].

Prior Probability

The estimated probability of an event before considering new evidence (e.g., disease prevalence in population).

Posterior Probability

The updated probability after incorporating new evidence (e.g., probability of disease after test result).

Sensitivity in Bayes' Theorem

The probability of a positive test given the presence of disease: P(B|A).

Specificity in Bayes' Theorem

The probability of a negative test given the absence of disease: P(B^c|A^c).