Data Science Masters

Self-study plan to achieve zero to mastery in data science.

Goals

Score top 20% in Kaggle competitions
Expert with different data types (text, image, audio, video)
Expert with different techniques (regression, SVM, deep learning, genetic algorithms, etc)
Familiar with modern tooling (python, pandas, scikit, R, tensorflow, apache spark, etc)
Expert with various problems (classification, search, clustering, prediction, recommendation, etc)

fundamentals (able to read and implement technical papers)
building at scale pipelines / architectures

Study Plan Overview

Module 0 - Highschool Math
Module 1 - College Math I (Calculus)
Module 2 - College Math II (Linear Algebra)
Module 3 - College Math III (Discrete Math)
Module 4 - College Math IV (Probability and Statistics)
Module 5 - Computation and Algorithms
Module 6 - Artificial Intelligence and Machine Learning
Module 7 - Deep Learning
Module 8 - Data Mining and Recommenders
Module 9 - NLP and Computer Vision
Module 10 - Cloud Computing Architectures / Data Center Engineering

It is recommended to look ahead so long as the general trend is that of finishing earlier blocks before later blocks.

Module 0 - Highschool math

Not everyone was lucky enough to have a good start with math growing up. The goal is to level the playing field - by the end of Block 0 you should feel like you went to a highschool with world class teachers and finished top of your math class.

Algebra

Required Reading

📚 The Joy of X

Module 1 - College Math I (Calculus)

Differential Equations

Supplementary Material

Module 2 - College Math II (Linear Algebra)

Required Reading

📚 Linear Algebra and its Applications
📚 Coding the Matrix

Supplementary Material

Module 3 - College Math III (Discrete Math)

Proofs

Proofs, Set theory, propositional logic, induction, invariants, state-machines

Graph Theory

Combinatorics

Series, Sequences, Recurrences

Number Theory

divisibility, gcd, modular arithmetic, bezouts identity, eulers totient theorem, fermats little theorem, "water-jug pouring", prime factorization, diaphantine equations, fundemental theorem of arithmetic

Discrete Probability

Abstract Algebra

Visual Group Theory

Supplementary Material

Module 4 - College Math IV (Probability and Statistics)

Module 5 - Computation and Algorithms

R Programming

Information Theory

https://www.khanacademy.org/computing/computer-science/informationtheory

Python and Computation and Data

Algorithms

Introduction
- Algorithmic thinking, peak finding
- Models of computation, Python cost model, document distance

Resources 📚 Introduction to Algorithms (CLRS)

Module 6 - Artificial Intelligence and Machine Learning

https://www.coursera.org/specializations/aml

Artificial Intelligence

Machine Learning

Machine Learning Specialization by University of Washington on Coursera

Module 7 - Deep Learning

Deep Learning by deeplearning.ai on Coursera

Neural Networks and Deep Learning
Improving Deep Neural Networks: Hyperparameter Tuning, Regularization, and Optimization
Structuring Machine Learning Projects
Convolutional Neural Networks
Sequence Models

Goals:

different activation functions (sigmoid/tanh/relu)
different cost functions
with and without bias units
classification and regression problems
text / binary / image / recommenders
batch vs stochastic
JS, Python, PHP, Matlab, TensorFlow, SciKitLearn
create visualizations and blog explanations
Audit best courses / books

tradziej / data-science-masters Goto Github PK