ISYE 6501 Exam 2025

Algorithm - -a step-by-step procedure designed to carry out a task Change Detection - -Identifying when a significant change has taken place Classification - -Separation of data into two or more categories Classifier - -A boundary that separates data into two or more categories Cluster - -A group of points that are identified as being similar or near each other Cluster Center - -In some clustering algorithms (k-means), the central point of a cluster center (CENTROID) Clustering - -Separation of points into similar or near groupings. Form of unsupervised learning. CUSUM - -change detection method that compares observed distribution mean with a threshold level of change. Short for Cumulative Sum (also cumsum) Deep Learning - -Neural Network model with many hidden layers Dimension - -A feature of the data points. EM Algorithm - -Expectation Maximization Algorithm. Algorithm with two steps (often iterated). 1. Finds the function for the expected likelihood of getting the response given current parameters. 2. Finds new parameter values that maximize probability Heuristic - -Algorithm that isn't guaranteed to find the optimal solution K-means - -Clustering algorithm (unsupervised), that works by defining k centroids and then mapping each point to the closest centroid. K-nearest neighbor (K-NN) - -Classification algorithm (supervised), that works by mapping a data point to the k closest neighbors to it. Kernel - -A type of function that computes the similarity between two inputs. thanks to what's sometimes known as the "kernel trick", non-linear classifiers can be found almost as easily as linear ones. Helps represent higher dimensional data sets. ISYE 6501 ISYE 6501 Learning - -Finding/discovering new patterns in data that can be applied to new data Machine - -Apparatus that can do something. in ml it often refers to the algorithm and the computer is run on. Margin - -for a single point, the distance between the point and the classification boundary; for a set of points the minimum distance between a point in the set and the classification boundary; Also called separation. Machine Learning - -Use of computer algorithms to learn and discover patterns or structure in data, without being programmed specifically for them. Misclassified - -To put a data point in the wrong category by a classifier Neural Network - -A machine learning model that itself is modeled after the workings of neurons in the brain. Supervised Learning - -Machine learning where the "correct" answer is known for each data point in the training set. Support Vector - -In SVM models, the point closest to the classifier, among those in the category. Support Vector Machine (SVM) - -Classification Algorithm (Supervised). Uses boundary to separate data into two or more classes Unsupervised Learning - -Machine learning where the "correct" answer is not known for the data points in the training set. Voronoi Diagram - -Graphical representation of splitting a plane into two or more special regions with one special point each, where each region's points are closer to the region's special point than to any other special point (Think K-means but visually represented) Accuracy - -Fraction of data points correctly classified by the model. (TP + TN)/(TP + TN + FP + FN) Confusion Matrix - -Visualization of classification model performance. Diagnostic odds ratio - -Ratio of the odds that a data point in a certain category is correctly classified by the model, to the odds that a data point is not in that category is incorrectly classified by the model; equal to (TP/FN)/(FP/TN) = (TP*TN)/(FP*FN) Fall out (FPR) - -Fraction of data points not in a certain category that are incorrectly classified by the model. (FP/TN+FP). ISYE 6501 ISYE 6501 False Negative - -When the model predicts the value is false but the value is actually true False Negative Rate (Miss Rate) - -Fraction of data points that are incorrectly classified by a model. FN/(FN+TP). False Positive - -When the model predicts that the value is true but the value is actually false False omission rate - -FN/(TN+FN) Hit rate (TPR, Sensitivity, Recall) ** - -TP/(TP+FN) Negative likelihood ratio - -Ratio of the fraction of data points in a certain category that are misclassified as not in the category, to the fraction of data points not in the category that are correctly classified as not being in the category. Equal to (1 sensitivity)/specificity = (FN/(FN+TP))/(TN/(TN+FP)) Positive likelihood ratio - -Ratio of the fraction of data points in a certain category that are correctly classified as being in that category, to the fraction of data points not in that category that are incorrectly classified as not being in that category. Equal to sensitivity/(1-specificity) = (TP/(TP+FN))/(TN/(TN+FP)) Negative Predictive Value (NPV) - -Fraction of data points classified as not in a certain category that are really not in that category. TN/(TN+FN) Positive Predictive Value (PPV, Precision) - -Fraction of data points classified in a certain category that TP/(TP+FP) Specificity (TNR) - -Fraction of data points not in a certain category that are correctly classified by the model (TN/TN+FP) True Negative - -Data point that a model correctly classifies as not being in a certain category. ("Negative" means the model classified it as not being in the category, and "True" means the model's classification is correct.) Sometimes abbreviated as "TN". True Positive - -Data point that a model correctly classifies as being in a certain category. ("Positive" means the model classified it as being in the category, and "True" means the model's classification is correct.) Sometimes abbreviated as "TP". Attribute/Covariate/Feature - -A characteristic or measurement - for example, a person's height or the color of a car. Generally interchangeable with "feature", and often with "covariate" or "predictor". In the standard tabular format, a column of data. Binary data - -Data that can take only two different values (true/false, 0/1, black/white, on/off, etc.). ISYE 6501 ISYE 6501 Box and whisker plot - -Graphical representation data showing the middle range of data (the "box"), reasonable ranges of variability ("whiskers"), and points (possible outliers) outside those ranges. Categorical data - -Data that classifies observations without quantitative meaning (for example, colors of cars) or where quantitative amounts are categorized (for example, "0-10, 11-20, ..."). Collective outlier - -A set of data points that is (uncommonly) different from others - for example, a missing heartbeat in an electrocardiogram; we don't know exactly which millisecond it should've happened in, but collectively there's a set of milliseconds that it's missing from. Expected but missed Contextual outlier - -A data point that is (uncommonly) far from other data points related to it - for example, in Atlanta, a 90-degree (Fahrenheit) day in winter is an outlier, but a 90-degree day in summer is not. Something way out there Data point - -Observation/record of (perhaps multiple) measurements for a single member of a population or data set. In the standard tabular format, a row of data. Detrending - -Removal of trend, such as a change in the mean over time, from time- series data. Eigenvalue - -Amount by which an eigenvector gets rescaled in a linear transformation. Eigenvector - -Non-zero vector that does not change direction when a linear transformation is applied to it, but only gets rescaled by the eigenvalue Imputation - -Inserting values where data is missing. Observation - -(1) A measurement of one attribute of a data point. (2) A measurement of all attributes of a data point (i.e., a full row of data). (3) The act of watching/measuring/recording something. Principal Component Analysis (PCA) - -Transformation of data into orthogonal dimensions that are ranked by variance. Point outlier - -A data point that is (uncommonly) far from other data points - for example, an outdoor temperature reading of 200 degrees Fahrenheit. Predictor - -A characteristic or measurement that is used to estimate ("predict")t he future value of something - for example, a person's height or the color of a car. A "feature" or "attribute"; in the standard tabular format, a column of data. ISYE 6501 ISYE 6501 Quantitative data - -Data that describes numerical amounts of something - for example, height and weight. Response - -A variable of interest that a model tries to estimate or predict. Scaling - -Shrinking or expanding, and moving, the range of data to fit exactly into a specific interval (for example, between 0 and 1, or between 100 and 800). Standardization - -Transforming data by subtracting the mean and then dividing by standard deviation, so that it has mean 0 and variance 1. Structured data - -Data that is highly organized, so it can be searched, queried, and analyzed easily - for example, a table with the name, age, and country of participants in this course. Time series data - -Data that records the same attribute/response at multiple points in time (often at equal time intervals). Unstructured data - -Data that is not very well organized for analysis - for example, a list of free responses to the question "What do you like about analytics?" Akaike information criterion (AIC) - -Model selection technique that trades off betw