-
Book Overview & Buying
-
Table Of Contents
15 Math Concepts Every Data Scientist Should Know
By :
15 Math Concepts Every Data Scientist Should Know
By:
Overview of this book
Data science combines the power of data with the rigor of scientific methodology, with mathematics providing the tools and frameworks for analysis, algorithm development, and deriving insights. As machine learning algorithms become increasingly complex, a solid grounding in math is crucial for data scientists. David Hoyle, with over 30 years of experience in statistical and mathematical modeling, brings unparalleled industrial expertise to this book, drawing from his work in building predictive models for the world's largest retailers.
Encompassing 15 crucial concepts, this book covers a spectrum of mathematical techniques to help you understand a vast range of data science algorithms and applications. Starting with essential foundational concepts, such as random variables and probability distributions, you’ll learn why data varies, and explore matrices and linear algebra to transform that data. Building upon this foundation, the book spans general intermediate concepts, such as model complexity and network analysis, as well as advanced concepts such as kernel-based learning and information theory. Each concept is illustrated with Python code snippets demonstrating their practical application to solve problems.
By the end of the book, you’ll have the confidence to apply key mathematical concepts to your data science challenges.
Table of Contents (21 chapters)
Preface
Chapter 1: Recap of Mathematical Notation and Terminology
Chapter 2: Random Variables and Probability Distributions
Chapter 3: Matrices and Linear Algebra
Chapter 4: Loss Functions and Optimization
Chapter 5: Probabilistic Modeling
Part 2: Intermediate Concepts
Chapter 6: Time Series and Forecasting
Chapter 7: Hypothesis Testing
Chapter 8: Model Complexity
Chapter 9: Function Decomposition
Chapter 10: Network Analysis
Part 3: Selected Advanced Concepts
Chapter 11: Dynamical Systems
Chapter 12: Kernel Methods
Chapter 13: Information Theory
Chapter 14: Non-Parametric Bayesian Methods
Chapter 15: Random Matrices
Index
is something we set, controlling the rate at which we get false positives. More correctly, the value of 
is the probability of falsely rejecting the null hypothesis. It is also called the Type-I error rate (i.e., the rate at which we make errors of Type-I). The term “false positive” tends to be used more when we assess accuracy after an event, when we are comparing to some known ground truth (i.e., what the correct decision or classification should have been). Type-I error is a term used more when the null hypothesis is falsely rejected and when discussing a hypothesis test a priori (e.g., when we are designing it). The difference between the terms “false positive” and “Type-I error” is subtle, and for the most part, we ignore it and use the 
