Skip to content

hanjie-jiang/bias-to-action

Repository files navigation

Bias to Action

Build intuition, then build models.

GitHub Pages MIT License PRs Welcome

A comprehensive, from-scratch learning resource covering algorithms, data structures, machine learning fundamentals, and modern transformer architectures — all in plain language.

Browse Documentation Learning Plan Star This Repo


Why This Repository?

Ever felt overwhelmed by scattered ML tutorials, confusing notation, and disconnected concepts? Bias to Action is different.

This is a complete, interconnected learning system that:

  • Starts from first principles — No assumed knowledge
  • Explains the "why" — Not just "what" or "how"
  • Connects everything — Math → Algorithms → ML → Transformers
  • Includes visualizations — LaTeX equations, algorithm diagrams, attention patterns
  • Provides practical implementations — Python code with complexity analysis
  • Covers cutting-edge topics — Attention mechanisms, BERT, GPT, modern transformers

Perfect for: Self-learners, bootcamp students, CS undergrads, ML engineers wanting to solidify fundamentals, and anyone preparing for technical interviews.


What Makes This Special?

Comprehensive & Interconnected

  • 85+ markdown files spanning mathematics, algorithms, ML, and deep learning
  • Cross-referenced content — Every topic links to related concepts
  • Progressive difficulty — Builds from fundamentals to state-of-the-art

Academic Rigor Meets Practical Clarity

  • MIT 6.006 Integration — Systematic algorithmic foundations
  • Plain language explanations — No unnecessary jargon
  • Step-by-step derivations — Follow the math from first principles
  • Real-world examples — LeetCode problems, practical applications

Modern Deep Learning

  • Complete Transformer Coverage — From attention basics to BERT vs GPT
  • Mathematical Deep Dives — Query-Key-Value, scaled dot-product, multi-head attention
  • Architecture Breakdowns — Encoder-decoder, positional encoding, layer normalization
  • Contemporary Models — GPT-3/4, RoBERTa, InstructGPT, ChatGPT

Beautiful, Searchable Interface

  • Modern UI — Responsive design with pastel theme
  • LaTeX rendering — Perfect mathematical equations via MathJax
  • Full-text search — Find anything instantly
  • Mobile-friendly — Learn anywhere, anytime

What's Inside

Mathematical Foundations

  • Linear Algebra for ML
  • Calculus & Gradient Descent
  • Asymptotic Analysis
  • Complexity Theory

Data Structures

  • Arrays & Linked Lists
  • Stacks (LIFO) & Queues (FIFO)
  • Hash Tables & Sets
  • Recursion Patterns

Algorithms

  • Binary Search & Variations
  • Sorting (Quicksort, Mergesort)
  • Two Pointers
  • Sliding Window

ML Fundamentals

  • Feature Engineering
  • Model Evaluation
  • Regularization (L1/L2)
  • Classical Algorithms

Probability & Stats

  • Bayes' Rule
  • Naive Bayes
  • Joint & Marginal Distributions
  • Markov Processes

Neural Networks

  • Perceptron Algorithm
  • Backpropagation
  • Deep Learning Fundamentals
  • N-gram Language Models

Attention & Transformers (Complete Modern Deep Learning)

Attention FundamentalsSelf-AttentionMulti-Head AttentionPositional EncodingTransformer ArchitectureBERT vs GPTModern Variants (GPT-3/4, RoBERTa, ChatGPT)


Quick Start

Option 1: Browse Online (Recommended)

Visit the live documentation — no setup required!

Option 2: Local Setup with Obsidian

Perfect for note-taking and personal customization:

# Clone the repository
git clone https://github.com/Hanjie-Jiang/bias-to-action.git
cd bias-to-action

# Open _notes/ folder in Obsidian
# Download Obsidian: https://obsidian.md/

Option 3: Run Locally with MkDocs

Build and serve the documentation yourself:

# Clone the repository
git clone https://github.com/Hanjie-Jiang/bias-to-action.git
cd bias-to-action

# Install dependencies
pip install -r requirements.txt

# Serve locally
mkdocs serve

# Open http://127.0.0.1:8000/bias-to-action/

Learning Path

I have designed a 20-week structured curriculum covering everything from fundamentals to transformers:

Phase 1: Foundations (Weeks 1-10)

  1. Weeks 1-2: Mathematical Foundations (Linear Algebra, Calculus)
  2. Weeks 3-4: Data Structures (Arrays, Linked Lists, Stacks, Queues)
  3. Weeks 5-6: Algorithms (Search, Sort, Complexity Analysis)
  4. Weeks 7-8: Probability & Statistics
  5. Weeks 9-10: Classical ML (Regression, Decision Trees, KNN)

Phase 2: Modern Deep Learning (Weeks 11-20)

  1. Weeks 11-12: Attention Mechanisms & Self-Attention
  2. Weeks 13-14: Multi-Head Attention & Positional Encoding
  3. Weeks 15-16: Transformer Architecture (Encoder-Decoder)
  4. Weeks 17-18: BERT & GPT (Masked vs Causal Language Modeling)
  5. Weeks 19-20: Advanced Topics & Integration Project

View Full Learning Plan


Key Features

Feature Description
Complete Coverage 85+ files covering math → algorithms → ML → transformers
Interconnected Every topic links to prerequisites and related concepts
Mathematical Rigor LaTeX equations with step-by-step derivations
Practical Code Python implementations with complexity analysis
LeetCode Integration Real interview problems with detailed solutions
Visual Learning Diagrams, attention visualizations, algorithm animations
Searchable Full-text search across all content
Responsive Works beautifully on desktop, tablet, and mobile

Contributing

Contributions are welcome and encouraged! Here's how you can help:

  • Report bugs or suggest improvements via Issues
  • Add new content — Fill in stub pages or expand existing topics
  • Fix typos — Even small improvements matter
  • Improve visualizations — Add diagrams or better examples
  • Share — Star the repo and tell others!

See our Contributing Guide for details.


Repository Structure

bias-to-action/
_notes/ # All markdown content
calculus_and_linear_algebra/ # Mathematical foundations
engineering_and_data_structure/ # Algorithms & data structures
ml_fundamentals/ # Classical ML
probability_and_markov/ # Statistics
neural_networks_and_deep_learning/ # Deep learning basics
attention_and_transformer_modern_neural_network/ # Transformers
assets/ # Images, CSS, resources
index.md # Landing page
mkdocs.yml # Site configuration
requirements.txt # Python dependencies
README.md # You are here!

Detailed structure: See Repository Structure below.


Highlights

Recent Additions (November 2025)

Attention & Transformers — Complete Modern Deep Learning

  • Attention Mechanism — Query-Key-Value framework from scratch
  • Self-Attention — Scaled dot-product with step-by-step math
  • Multi-Head Attention — Why multiple heads? Parallel representation learning
  • Positional Encoding — Sinusoidal vs learned embeddings
  • Transformer Architecture — Full encoder-decoder breakdown
  • BERT vs GPT — Bidirectional vs causal language modeling
  • Modern Variants — GPT-3/4, InstructGPT, ChatGPT, RoBERTa

Data Structures & Algorithms Mastery

  • Stacks & Queues — LIFO/FIFO operations with real-world applications
  • Monotonic Stack — Next greater/smaller element patterns
  • Quicksort Deep Dive — Partitioning visualizations step-by-step
  • Binary Search Variations — Peak finding, rotated arrays

Stats

  • 85+ Markdown Files — Comprehensive coverage
  • 8 Major Topics — Math, algorithms, ML, transformers
  • 20-Week Curriculum — Structured learning path
  • 100% Open Source — Free forever
  • Active Development — Regular updates and improvements

Show Your Support

If you found this helpful:

  • Star this repository to help others discover it
  • Fork it to customize for your own learning
  • Share it with friends, classmates, or study groups
  • Provide feedback via Issues or Discussions

Every star motivates us to keep improving!


License

This project is licensed under the MIT License — see the LICENSE file for details.

Free to use, modify, and share. Just keep the attribution!


Links


Contact

Questions? Suggestions? Feel free to open an Issue or start a Discussion.


Made with for learners, by learners

Build intuition, then build models.

Star this repo Fork it Start Learning


Repository Structure (Detailed)

Click to expand full directory structure
_notes/
index.md # Main landing page
Foundational knowledge plan.md # 20-week learning roadmap
Information_Theory.md # Information theory concepts
Integration_and_Project.md # Integration projects

assets/ # Website resources
images/ # General image resources
ml_fundamentals/ # ML fundamentals resources
styles/ # Custom CSS
hero.css # Hero section styling
layout.css # Main layout styling

calculus_and_linear_algebra/ # Mathematical Foundations
Calculus_and_Linear_Algebra_Overview.md
Linear_Algebra_for_ML.md # Vectors, matrices, operations
Calculus_and_Gradient_Descent.md # Optimization methods
Asymptotic_Analysis_Theory.md # Complexity theory

engineering_and_data_structure/ # Programming & Data Structures
Overview/
Engineering_and_Data_Structure_Overview.md
Data_Structures/
Arrays/
Arrays_Overview.md # Arrays vs linked lists
Dynamic_Arrays.md # Resizable arrays
Array_Problems.md # LeetCode problems
Linked_Lists/
Linked_Lists_Overview.md # Singly, doubly, circular
Linked_List_Implementation.md # Python implementation
Linked_List_Problems.md # Two pointers, reversal
Stacks/
Stacks_Overview.md # LIFO operations
Stack_Implementation.md # Array vs linked list
Stack_Problems.md # Valid parentheses, DFS
Queues/
Queues_Overview.md # FIFO operations
Queue_Implementation.md # Python deque
Queue_Problems.md # BFS, sliding window
Hash_Tables/
Hash_Tables_Overview.md
Hash_Functions_and_Collisions.md
Python_Dictionaries.md
Python_Dictionary_Operations.md
Python_Sets.md
Python_Set_Operations.md
Hash_Table_Problems.md
Recursion/
Recursion_Fundamentals.md
Recursive_Algorithms.md
Recursion_vs_Iteration.md
Common_Recursive_Patterns.md
Algorithms/
Search_Algorithms/
Search_Algorithms_Overview.md
Binary_Search_Fundamentals.md
Binary_Search_Variations.md
Search_Problems.md # Peak finding
Sorting_Algorithms/
Sorting_Algorithms_Overview.md # Quicksort deep dive
Sorting_Problems.md # LeetCode problems
Algorithmic_Patterns/
Two_Pointers/
Two_Pointers_Overview.md
Opposite_Direction_Pointers.md
Same_Direction_Pointers.md
Two_Pointers_Problems.md
Sliding_Window/
Sliding_Window_Overview.md
Fixed_Size_Window.md
Variable_Size_Window.md
Sliding_Window_Problems.md
Problem_Solving/
Set_Dictionary_Problems/
Array_Intersection.md
Non_Repeating_Elements.md
Unique_Elements.md
Anagram_Pairs.md
String_Problems/
Unique_Strings.md
Resources/
MIT_6006_Integration_Templates.md # MIT algorithm templates
Common_Patterns.md
Time_Complexity_Guide.md
Interview_Strategies.md

ml_fundamentals/ # Machine Learning Fundamentals
ML_Fundamentals_Overview.md
feature_engineering/
categorical_encoding.md
data_types_and_normalization.md
feature_crosses.md
model_evaluation/
evaluation_methods.md
metrics_and_validation.md
hyperparameter_tuning.md
regularization/
overfitting_underfitting.md
l1_l2_regularization.md
early_stopping.md
classical_algorithms/
linear_regression.md
logistic_regression.md
decision_trees.md
unsupervised_learning/
k_nearest_neighbors.md
k_means_clustering.md

language_model/ # Natural Language Processing
Ngram_Language_Modeling.md # N-gram models
resources/
Happy-LLM-v1.0.pdf # Reference materials

neural_networks_and_deep_learning/ # Deep Learning
Neural_Networks_and_Deep_Learning_Overview.md
neural_networks_sections/
Introduction_to_Perceptron_Algorithm.md

probability_and_markov/ # Probability & Statistics
Probability_and_Markov_Overview.md
probability_and_markov_sections/
conditional_probability_and_bayes_rule.md
joint_and_marginal_distributions.md
naive_bayes_and_gaussian_naive_bayes.md
resources/
conditional_probability.png

attention_and_transformer_modern_neural_network/ # Modern Deep Learning
Attention_and_Transformers_Overview.md # Complete overview
attention_fundamentals/
Attention_Mechanism_Overview.md # Q-K-V framework
Attention_Math.md # Scaled dot-product
self_attention/
Self_Attention_Overview.md # Self-attention mechanism
multi_head_attention/
Multi_Head_Attention_Overview.md # Multi-head attention
positional_encoding/
Positional_Encoding_Overview.md # Position embeddings
transformer_architecture/
Transformer_Architecture_Overview.md # Encoder-decoder
transformer_variants/
BERT_and_GPT_Overview.md # BERT vs GPT

javascripts/ # Website functionality
mathjax.js # LaTeX rendering
floating-nav.js # Navigation enhancements

Changelog

View update history

version 2025-11-16

  • Attention & Transformers Section: Added comprehensive modern deep learning content covering attention mechanisms, transformers, BERT, and GPT
  • Attention Fundamentals: Complete Query-Key-Value framework with scaled dot-product attention mathematics and intuitive explanations
  • Self-Attention Mechanism: Detailed coverage of self-attention with step-by-step computations and practical applications
  • Multi-Head Attention: Comprehensive overview explaining parallel attention heads, dimensionality, and representation subspaces
  • Positional Encoding: Sinusoidal and learned position embeddings explaining how transformers capture sequential information
  • Transformer Architecture: Full encoder-decoder architecture with feed-forward networks, layer normalization, and residual connections
  • BERT vs GPT: Complete comparison of encoder-only (BERT, masked language modeling) vs decoder-only (GPT, causal language modeling) architectures
  • Modern Variants: Coverage of RoBERTa, GPT-2/3/4, InstructGPT, ChatGPT, and contemporary transformer developments
  • Navigation Integration: Updated front page hero section, sidebar navigation, and learning plan with Weeks 11-20 for transformer content
  • Repository Cleanup: Added comprehensive .gitignore, removed build artifacts and OS metadata, cleaned nested directory structure

version 2025-10-21

  • Queue Implementation Mastery: Enhanced Queues Overview with comprehensive Python deque implementation and performance analysis
  • FIFO vs LIFO Comparison: Added detailed comparison tables between queues, stacks, and priority queues with complexity analysis
  • Data Structure Selection Guide: Added decision matrix for when to use queues vs other data structures with practical examples

version 2025-10-20

  • Monotonic Stack Problems: Added Daily Temperatures (LeetCode #739) with comprehensive solution using decreasing monotonic stack
  • Stack Problem Collection: Enhanced Stack Problems section with detailed explanations of monotonic stack techniques
  • Algorithm Optimization: Demonstrated how monotonic stacks improve from O(n²) brute force to O(n) optimal solutions

version 2025-10-13

  • Complete Data Structures Suite: Added comprehensive coverage of linked lists, stacks, and queues
  • Linear Data Structures Mastery: Detailed theory, implementations, and problem patterns for LIFO and FIFO operations
  • Navigation Integration: Updated MkDocs navigation, front page cards, and cross-references

version 2025-10-09

  • Sorting Algorithms Mastery: Completed comprehensive sorting section with detailed quicksort visualizations
  • Search & Sort Integration: Finalized both search and sorting algorithms with theoretical depth

Earlier versions

See CHANGELOG.md for complete history.

About

a git repo storing ML notes that contains basic concept and easy-to-understand resources that one can find online

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Contributors