Analytics Techniques — Animated

Interactive visualizations covering every technique in MADT 6004 — from descriptive statistics through inference, regression, time series, classification, evaluation, text analytics, and clustering. Each card opens a standalone interactive demo.

Chapter 2 Exploratory Data Analysis Summarize and visualize one variable at a time.
01Central TendencyMean, median, mode — and why an outlier wrecks one but not the others. 02Percentiles & QuartilesSlice a sorted distribution — quartiles, deciles, percentiles, box plots. 03Variance & Std DevSpread measured as squared deviations from the mean. 04DistributionsNormal curve, skewness, bimodal, uniform — with live sampling. 13Box Plot & IQR Outliers5-number summary + Tukey's k·IQR rule for outliers.
Chapter 3 Group Comparison Are two or more groups truly different, or just noise?
07Hypothesis TestingPopulation, sample, z-statistic, p-value, tail logic. 08Confidence IntervalsWhat "95% confidence" really means — through repeated sampling. 14t-test (Indep & Paired)Two groups: separate samples vs same subjects before/after. 15One-way ANOVAThree+ groups. F = between-variance / within-variance. 16Chi² Goodness of FitDo observed counts match expected? Σ(O−E)²/E.
Chapter 4 Correlation & Association How two variables move together — for numbers and for categories.
05Pearson Correlationr from −1 to +1 — linear association visualized as cloud shape. 17Spearman RankRank-based correlation — catches monotonic, not just linear. 18Chi² IndependenceAre two categorical variables related? Contingency table version.
Chapter 5 Regression Predict a continuous y. Choose models that don't overfit.
06Linear RegressionFit a line by minimizing squared residuals (OLS). 10Gradient DescentHow models learn — steps downhill on a loss surface. 19Multiple Regression & R²Multiple predictors + Adjusted R² penalty for useless features. 20Bias-Variance TradeoffUnderfit (high bias) ↔ overfit (high variance) via polynomial degree. 21Ridge vs LassoL2 vs L1 regularization — shrink coefficients or zero them out.
Chapter 5 Time Series Forecast values that depend on their own past.
22TS DecompositionPull a series apart into trend + seasonality + residual. 23Moving Avg & SmoothingMA, Simple Exp Smoothing, Holt — three forecasters compared. 24ARIMA(p, d, q)AR + Integration + MA — the workhorse of classical forecasting.
Chapter 6 Classification Predict a class label from features.
11Decision TreeRecursive splitting of feature space — and the tree it builds. 12Neural NetworkInputs → hidden → output, trained by gradient descent. 25Logistic RegressionSigmoid σ(z) maps log-odds to probabilities. 26K-Nearest NeighborsClassify a new point by its k closest training neighbors. 27Random ForestBag many trees on bootstrap samples — average smooths noise.
Chapter 7 Classification Model Evaluation How well did your classifier actually do?
28Confusion Matrix & MetricsTP, FP, FN, TN → accuracy, precision, recall, F1. 29ROC Curve & AUCThreshold sweep, TPR vs FPR, area under curve.
Chapter 7 Text Analytics Turn raw text into countable features.
30Tokenization & N-gramsPipeline: tokenize → lowercase → stopwords → stem → n-grams. 31TF-IDFScore = how unique a word is to a specific document.
Chapter 8 Similarity, Recommendation & Clustering Group like with like; recommend new items to users.
09K-Means ClusteringIteratively assign and recompute centroids until they settle. 32Distance MeasuresEuclidean, Manhattan, Chebyshev, Cosine, Jaccard. 33Hierarchical ClusteringMerge nearest pair repeatedly — dendrogram + cut at any height. 34Collaborative Filtering"Users like you also rated…" — neighbor-based recommendation.