From raw chaos to ML-ready features. Pipelines, ingestion, EDA, cleaning, outliers, encoding, feature engineering, selection, dimensionality reduction, time series, imbalance, augmentation, splits, cross-validation, and drift — 15 lessons covering the 80% of ML work that decides whether your model succeeds.
From raw sources to ML-ready datasets — understand the full data journey.
APIs, databases, files, streams — pull data out of any source reliably, with the right file format and sampling strategy.
Discover patterns, outliers, and stories hidden in your data.
Handle missing values, outliers, and duplicates like a pro.
Z-score, IQR, Mahalanobis, Isolation Forest, LOF, DBSCAN — the full anomaly toolkit, and when to drop, cap, transform, or keep.
One-hot, ordinal, target, frequency, hashing, embeddings — turn categories into numbers without lying or leaking.
Transform raw data into features that make models powerful.
Filter, wrapper, and embedded methods — drop the noise and keep only the features that move the needle.
PCA, t-SNE, UMAP — compress high-dimensional data while keeping the structure that matters.
Resampling, lags, rolling windows, cyclical encoding, stationarity — preparing time-stamped data without leaking the future.
SMOTE, class weighting, and the right metrics when your data is 99% one class.
Generate more training data from what you already have.
Split data correctly and avoid the silent killer of ML projects.
K-fold, stratified, group, time-series, nested CV — estimate generalization without fooling yourself.
KS test, PSI, Great Expectations, Pandera — catch covariate shift, label shift, and concept drift before they break production.
9 interactive labs — hands-on exercises for this track
A photo is just a grid of numbers — reshape, stack, and transform data like a pro
Your boss wants a report by Friday — pick the right chart to tell the story your data is hiding
Follow a patient's medical data from raw spreadsheet to trained model that predicts heart disease risk
Before training any model, data scientists spend 80% of their time here — exploring distributions, correlations, and outliers.
Watch k-fold cross-validation cycle through folds as the test set, with per-fold accuracy bars and mean ± std vs a single train/test split.
Side-by-side training pipelines expose preprocessing contamination. Watch a 99% test score collapse to 51% when the leaky model meets fresh data.
Toggle StandardScaler and MinMaxScaler on a k-NN classifier. Watch the decision boundary swing from horizontal (income dominates) to diagonal (features on equal footing).
Watch SMOTE manufacture synthetic minority samples one at a time — anchor, k-NN lines, interpolate, drop. See the logistic boundary shift as balance is restored.
Label / one-hot / target / embedding on the same column. Cardinality slider (5 / 100 / 1000) — watch one-hot collapse, embedding win.
375 questions across 15 modules — check how well you understood this track.