Taxonomy Distribution Report¶

Overview¶

• Total Concepts: 200
• Number of Taxonomies: 14
• Average Concepts per Taxonomy: 14.3

Distribution Summary¶

Visual Distribution¶

NN     █████████  37 ( 18.5%)
FOUND  ███████  31 ( 15.5%)
CNN    █████  20 ( 10.0%)
EVAL   ████  19 (  9.5%)
SVM    ████  16 (  8.0%)
OPT    ████  16 (  8.0%)
TREE   ███  12 (  6.0%)
CLUST  ███  12 (  6.0%)
KNN    ██  11 (  5.5%)
LOGREG ██   9 (  4.5%)
PREP   █   7 (  3.5%)
REG    █   5 (  2.5%)
TL     █   4 (  2.0%)
MISC      1 (  0.5%)

Balance Analysis¶

✅ No Over-Represented Categories¶

All categories are under the 30% threshold. Good balance!

ℹ️ Under-Represented Categories (<3%)¶

• Regularization (REG): 5 concepts (2.5%)
• Note: Small categories are acceptable for specialized topics
• Transfer Learning (TL): 4 concepts (2.0%)
• Note: Small categories are acceptable for specialized topics
• Miscellaneous (MISC): 1 concepts (0.5%)
• Note: Small categories are acceptable for specialized topics

Category Details¶

Neural Networks (NN)¶

Count: 37 concepts (18.5%)

Concepts:

• 1. Neural Network
• 1. Artificial Neuron
• 1. Perceptron
• 1. Activation Function
• 1. ReLU
• 1. Tanh
• 1. Leaky ReLU
• 1. Weights
• 1. Bias
• 1. Forward Propagation
• 1. Backpropagation
• 1. Gradient Descent
• 1. Stochastic Gradient Descent
• 1. Mini-Batch Gradient Descent
• 1. Learning Rate
• ...and 22 more

Foundation Concepts (FOUND)¶

Count: 31 concepts (15.5%)

Concepts:

• 1. Machine Learning
• 1. Supervised Learning
• 1. Unsupervised Learning
• 1. Classification
• 1. Regression
• 1. Training Data
• 1. Test Data
• 1. Validation Data
• 1. Feature
• 1. Label
• 1. Instance
• 1. Feature Vector
• 1. Model
• 1. Algorithm
• 1. Hyperparameter
• ...and 16 more

Convolutional Networks (CNN)¶

Count: 20 concepts (10.0%)

Concepts:

• 1. Convolutional Neural Network
• 1. Convolution Operation
• 1. Filter
• 1. Stride
• 1. Padding
• 1. Valid Padding
• 1. Same Padding
• 1. Receptive Field
• 1. Max Pooling
• 1. Average Pooling
• 1. Spatial Hierarchies
• 1. Translation Invariance
• 1. Local Connectivity
• 1. Weight Sharing
• 1. CNN Architecture
• ...and 5 more

Evaluation Metrics (EVAL)¶

Count: 19 concepts (9.5%)

Concepts:

• 1. Training Error
• 1. Test Error
• 1. Generalization
• 1. Stratified Sampling
• 1. Holdout Method
• 1. Confusion Matrix
• 1. True Positive
• 1. False Positive
• 1. True Negative
• 1. False Negative
• 1. Accuracy
• 1. Precision
• 1. Recall
• 1. F1 Score
• 1. ROC Curve
• ...and 4 more

Support Vector Machines (SVM)¶

Count: 16 concepts (8.0%)

Concepts:

• 1. Support Vector Machine
• 1. Hyperplane
• 1. Margin
• 1. Support Vectors
• 1. Margin Maximization
• 1. Hard Margin SVM
• 1. Soft Margin SVM
• 1. Slack Variables
• 1. Kernel Trick
• 1. Linear Kernel
• 1. Polynomial Kernel
• 1. Radial Basis Function
• 1. Gaussian Kernel
• 1. Dual Formulation
• 1. Primal Formulation
• ...and 1 more

Optimization (OPT)¶

Count: 16 concepts (8.0%)

Concepts:

• 1. Computational Complexity
• 1. Time Complexity
• 1. Space Complexity
• 1. Scalability
• 1. Online Learning
• 1. Optimizer
• 1. Adam Optimizer
• 1. RMSprop
• 1. Momentum
• 1. Nesterov Momentum
• 1. Gradient Clipping
• 1. Dropout
• 1. Early Stopping
• 1. Grid Search
• 1. Random Search
• ...and 1 more

Decision Trees (TREE)¶

Count: 12 concepts (6.0%)

Concepts:

• 1. Decision Tree
• 1. Tree Node
• 1. Leaf Node
• 1. Splitting Criterion
• 1. Entropy
• 1. Information Gain
• 1. Gini Impurity
• 1. Pruning
• 1. Overfitting
• 1. Underfitting
• 1. Tree Depth
• 1. Cross-Entropy Loss

Clustering (CLUST)¶

Count: 12 concepts (6.0%)

Concepts:

• 1. K-Means Clustering
• 1. Centroid
• 1. Cluster Assignment
• 1. Cluster Update
• 1. K-Means Initialization
• 1. Random Initialization
• 1. K-Means++ Initialization
• 1. Elbow Method
• 1. Silhouette Score
• 1. Within-Cluster Variance
• 1. Convergence Criteria
• 1. Inertia

K-Nearest Neighbors (KNN)¶

Count: 11 concepts (5.5%)

Concepts:

• 1. K-Nearest Neighbors
• 1. Distance Metric
• 1. Euclidean Distance
• 1. Manhattan Distance
• 1. K Selection
• 1. Decision Boundary
• 1. Voronoi Diagram
• 1. Curse of Dimensionality
• 1. KNN for Classification
• 1. KNN for Regression
• 1. Lazy Learning

Logistic Regression (LOGREG)¶

Count: 9 concepts (4.5%)

Concepts:

• 1. Sigmoid Function
• 1. Log-Loss
• 1. Binary Classification
• 1. Multiclass Classification
• 1. Maximum Likelihood
• 1. One-vs-All
• 1. One-vs-One
• 1. Softmax Function
• 1. Sigmoid Activation

Data Preprocessing (PREP)¶

Count: 7 concepts (3.5%)

Concepts:

• 1. Normalization
• 1. Standardization
• 1. Min-Max Scaling
• 1. Z-Score Normalization
• 1. One-Hot Encoding
• 1. Dimensionality Reduction
• 1. Data Augmentation

Regularization (REG)¶

Count: 5 concepts (2.5%)

Concepts:

• 1. Regularization
• 1. L1 Regularization
• 1. L2 Regularization
• 1. Ridge Regression
• 1. Lasso Regression

Transfer Learning (TL)¶

Count: 4 concepts (2.0%)

Concepts:

• 1. Transfer Learning
• 1. Fine-Tuning
• 1. Domain Adaptation
• 1. ImageNet

Miscellaneous (MISC)¶

Count: 1 concepts (0.5%)

Concepts:

• 1. Logistic Regression

Recommendations¶

• ✅ Good balance: Categories are reasonably distributed (spread: 18.0%)
• ✅ MISC category minimal: Good categorization specificity

Educational Use Recommendations¶

• Use taxonomy categories for color-coding in graph visualizations
• Design curriculum modules based on taxonomy groupings
• Create filtered views for focused learning paths
• Use categories for assessment organization
• Enable navigation by topic area in interactive tools

Report generated by learning-graph-reports/taxonomy_distribution.py