Algorithm Workspace

Centralized catalog for interactive, zero-dependency model implementations built from scratch.

Total Engines14
Paradigm Split8 Supervised | 2 Unsupervised
⌕
All EnginesSupervisedUnsupervised
Supervised Gradient Descent / Chain Rule

Neural Network / Embedding Engine

Multi-layer matrix lookup architecture updating weights and processing structural vector transformations live.

Train ComplexityO(E · V · D)
Inference SpeedFast
ClassificationOpen Canvas →
Unsupervised Euclidean Space L2 Norm

k-Means Engine

Iterative partition engine moving k cluster centroids to center points until cluster boundary stabilization.

Train ComplexityO(I · K · N · M)
Inference SpeedModerate
ClusteringOpen Canvas →
Supervised Conditional Probability

Naive Bayes Classifier

Log-likelihood probabilistic calculator utilizing strict feature independence assumptions for instant training.

Train ComplexityO(N · D)
Inference SpeedExtremely Fast
ClassificationOpen Canvas →
Supervised Instance Distance Matching

k-Nearest Neighbors

Lazy non-parametric algorithm mapping immediate neighborhood density matrices without an explicit training step.

Train ComplexityO(1)
Inference SpeedSlow O(N · D)
ClassificationOpen Canvas →
Supervised Ordinary Least Squares

Linear Regression Engine

Fits a continuous optimized hyperplane across scalar metrics by minimizing residual sums of squares.

Train ComplexityO(D³ + N·D²)
Inference SpeedInstant O(D)
RegressionOpen Canvas →
Supervised Sigmoid / Cross-Entropy

Logistic Regression

Maps binary or multinomial classes by driving dot products through a logistic function to bound outputs between 0 and 1.

Train ComplexityO(N · D)
Inference SpeedInstant O(D)
ClassificationOpen Canvas →
Supervised Hinge Loss / Convex Opt

Support Vector Machine

Maximizes the geometric margin between boundary vectors. Includes structural kernel options for non-linear boundary maps.

Train ComplexityO(N² · D) to O(N³ · D)
Inference SpeedO(S · D)
ClassificationOpen Canvas →
Supervised Shannon Entropy / Gini Impurity

Decision Tree (ID3/CART)

Greedy recursive feature splitting engine built on information gain metrics to construct explicit programmatic parsing branches.

Train ComplexityO(D · N log N)
Inference SpeedFast O(Tree Depth)
ClassificationOpen Canvas →
Supervised Bootstrap Aggregation (Bagging)

Random Forest Ensemble

Constructs an uncorrelated forest of structural decision trees, averaging node parsing to drop model variance errors.

Train ComplexityO(T · D · N log N)
Inference SpeedO(T · Tree Depth)
ClassificationOpen Canvas →
Unsupervised Covariance Eigenvalues / SVD

Principal Component Analysis (PCA)

Orthogonal linear transform mapping data covariance vectors onto high-variance directions to condense dimensional profiles.

Train ComplexityO(D³ + N · D²)
Inference SpeedO(D · K)
Dimensionality ReductionOpen Canvas →
Unsupervised Covariance Eigenvalues / SVD

Peceptron Classifier

Orthogonal linear transform mapping data covariance vectors onto high-variance directions to condense dimensional profiles.

Train ComplexityO(D³ + N · D²)
Inference SpeedO(D · K)
Dimensionality ReductionOpen Canvas →
Unsupervised Covariance Eigenvalues / SVD

Polynomial Regression (PCA)

Orthogonal linear transform mapping data covariance vectors onto high-variance directions to condense dimensional profiles.

Train ComplexityO(D³ + N · D²)
Inference SpeedO(D · K)
Dimensionality ReductionOpen Canvas →
Supervised Covariance Eigenvalues / SVD

Linear SVM

Orthogonal linear transform mapping data covariance vectors onto high-variance directions to condense dimensional profiles.

Train ComplexityO(D³ + N · D²)
Inference SpeedO(D · K)
Dimensionality ReductionOpen Canvas →
Supervised Covariance Eigenvalues / SVD

Train SVM

Orthogonal linear transform mapping data covariance vectors onto high-variance directions to condense dimensional profiles.

Train ComplexityO(D³ + N · D²)
Inference SpeedO(D · K)
Dimensionality ReductionOpen Canvas →