Master Statistics & Machine Learning: Intuition, Math, Code

Master Statistics & Machine Learning: Intuition, Math, Code
Last updated 10/2022
MP4 | Video: h264, 1920×1080 | Audio: AAC, 44.1 KHz

A rigorous and engaging deep-dive into statistics and machine-learning, with hands-on applications in Python and MATLAB.

What you’ll learn
Inferential statistics
T-tests, correlation, ANOVA, regression, clustering
The math behind the "black box" statistical methods
How to implement statistical methods in code
How to interpret statistics correctly and avoid common misunderstandings
Coding techniques in Python and MATLAB/Octave
Machine learning methods like clustering, predictive analysis, classification, and data cleaning
Requirements
Good work ethic and motivation to learn.
Previous background in statistics or machine learning is not necessary.
Python -OR- MATLAB with the Statistics toolbox (or Octave).
Some coding familiarity for the optional code exercises.
No textbooks necessary! All materials are provided inside the course.
Description
Overview

Section 1: Introductions

Lecture 1 [Important] Getting the most out of this course

Lecture 2 About using MATLAB or Python

Lecture 3 Statistics guessing game!

Lecture 4 Using the Q&A forum

Lecture 5 (optional) Entering time-stamped notes in the Udemy video player

Section 2: Math prerequisites

Lecture 6 Should you memorize statistical formulas?

Lecture 7 Arithmetic and exponents

Lecture 8 Scientific notation

Lecture 9 Summation notation

Lecture 10 Absolute value

Lecture 11 Natural exponent and logarithm

Lecture 12 The logistic function

Lecture 13 Rank and tied-rank

Section 3: IMPORTANT: Download course materials

Lecture 14 Download materials for the entire course!

Section 4: What are (is?) data?

Lecture 15 Is "data" singular or plural?!?!!?!

Lecture 16 Where do data come from and what do they mean?

Lecture 17 Types of data: categorical, numerical, etc

Lecture 19 Sample vs. population data

Lecture 20 Samples, case reports, and anecdotes

Lecture 21 The ethics of making up data

Section 5: Visualizing data

Lecture 22 Bar plots

Lecture 23 Code: bar plots

Lecture 24 Box-and-whisker plots

Lecture 25 Code: box plots

Lecture 26 "Unsupervised learning": Boxplots of normal and uniform noise

Lecture 27 Histograms

Lecture 28 Code: histograms

Lecture 29 "Unsupervised learning": Histogram proportion

Lecture 32 When to use lines instead of bars

Lecture 33 Linear vs. logarithmic axis scaling

Lecture 34 Code: line plots

Lecture 35 "Unsupervised learning": log-scaled plots

Section 6: Descriptive statistics

Lecture 36 Descriptive vs. inferential statistics

Lecture 37 Accuracy, precision, resolution

Lecture 38 Data distributions

Lecture 39 Code: data from different distributions

Lecture 40 "Unsupervised learning": histograms of distributions

Lecture 41 The beauty and simplicity of Normal

Lecture 42 Measures of central tendency (mean)

Lecture 43 Measures of central tendency (median, mode)

Lecture 44 Code: computing central tendency

Lecture 45 "Unsupervised learning": central tendencies with outliers

Lecture 47 Code: Computing dispersion

Lecture 49 Code: IQR

Lecture 50 QQ plots

Lecture 51 Code: QQ plots

Lecture 52 Statistical "moments"

Lecture 54 Code: Histogram bins

Lecture 55 Violin plots

Lecture 56 Code: violin plots

Lecture 57 "Unsupervised learning": asymmetric violin plots

Lecture 58 Shannon entropy

Lecture 59 Code: entropy

Lecture 60 "Unsupervised learning": entropy and number of bins

Section 7: Data normalizations and outliers

Lecture 61 Garbage in, garbage out (GIGO)

Lecture 62 Z-score standardization

Lecture 63 Code: z-score

Lecture 64 Min-max scaling

Lecture 65 Code: min-max scaling

Lecture 66 "Unsupervised learning": Invert the min-max scaling

Lecture 67 What are outliers and why are they dangerous?

Lecture 68 Removing outliers: z-score method

Lecture 69 The modified z-score method

Lecture 70 Code: z-score for outlier removal

Lecture 71 "Unsupervised learning": z vs. modified-z

Lecture 73 Code: Euclidean distance for outlier removal

Lecture 74 Removing outliers by data trimming

Lecture 75 Code: Data trimming to remove outliers

Lecture 76 Non-parametric solutions to outliers

Lecture 77 Nonlinear data transformations

Lecture 78 An outlier lecture on personal accountability

Section 8: Probability theory

Lecture 79 What is probability?

Lecture 80 Probability vs. proportion

Lecture 81 Computing probabilities

Lecture 82 Code: compute probabilities

Lecture 83 Probability and odds

Lecture 84 "Unsupervised learning": probabilities of odds-space

Lecture 85 Probability mass vs. density

Lecture 86 Code: compute probability mass functions

Lecture 87 Cumulative distribution functions

Lecture 88 Code: cdfs and pdfs

Lecture 90 Creating sample estimate distributions

Lecture 91 Monte Carlo sampling

Lecture 94 Expected value

Lecture 95 Conditional probability

Lecture 96 Code: conditional probabilities

Lecture 97 Tree diagrams for conditional probabilities

Lecture 98 The Law of Large Numbers

Lecture 99 Code: Law of Large Numbers in action
Lecture 100 The Central Limit Theorem

Lecture 101 Code: the CLT in action
Lecture 102 "Unsupervised learning": Averaging pairs of numbers

Section 9: Hypothesis testing

Lecture 103 IVs, DVs, models, and other stats lingo

Lecture 104 What is an hypothesis and how do you specify one?

Lecture 105 Sample distributions under null and alternative hypotheses

Lecture 106 P-values: definition, tails, and misinterpretations

Lecture 107 P-z combinations that you should memorize

Lecture 108 Degrees of freedom

Lecture 109 Type 1 and Type 2 errors

Lecture 110 Parametric vs. non-parametric tests

Lecture 111 Multiple comparisons and Bonferroni correction

Lecture 112 Statistical vs. theoretical vs. clinical significance

Lecture 113 Cross-validation

Lecture 114 Statistical significance vs. classification accuracy

Section 10: The t-test family

Lecture 115 Purpose and interpretation of the t-test

Lecture 116 One-sample t-test

Lecture 117 Code: One-sample t-test

Lecture 119 Two-samples t-test

Lecture 120 Code: Two-samples t-test

Lecture 121 "Unsupervised learning": Importance of N for t-test

Lecture 122 Wilcoxon signed-rank (nonparametric t-test)

Lecture 123 Code: Signed-rank test

Lecture 124 Mann-Whitney U test (nonparametric t-test)

Lecture 125 Code: Mann-Whitney U test

Lecture 126 Permutation testing for t-test significance

Lecture 127 Code: permutation testing

Lecture 128 "Unsupervised learning": How many permutations?

Section 11: Confidence intervals on parameters

Lecture 129 What are confidence intervals and why do we need them?

Lecture 130 Computing confidence intervals via formula

Lecture 131 Code: compute confidence intervals by formula

Lecture 132 Confidence intervals via bootstrapping (resampling)

Lecture 133 Code: bootstrapping confidence intervals

Lecture 135 Misconceptions about confidence intervals

Section 12: Correlation

Lecture 136 Motivation and description of correlation

Lecture 138 Code: correlation coefficient

Lecture 139 Code: Simulate data with specified correlation

Lecture 140 Correlation matrix

Lecture 141 Code: correlation matrix

Lecture 142 "Unsupervised learning": average correlation matrices

Lecture 146 The problem with Pearson

Lecture 147 Nonparametric correlation: Spearman rank

Lecture 148 Fisher-Z transformation for correlations

Lecture 149 Code: Spearman correlation and Fisher-Z

Lecture 150 "Unsupervised learning": Spearman correlation

Lecture 151 "Unsupervised learning": confidence interval on correlation

Lecture 152 Kendall’s correlation for ordinal data

Lecture 153 Code: Kendall correlation

Lecture 154 "Unsupervised learning": Does Kendall vs. Pearson matter?

Lecture 155 The subgroups correlation paradox

Lecture 156 Cosine similarity

Lecture 157 Code: Cosine similarity vs. Pearson correlation

Lecture 160 Sum of squares

Lecture 161 The F-test and the ANOVA table

Lecture 162 The omnibus F-test and post-hoc comparisons

Lecture 163 The two-way ANOVA

Lecture 164 One-way ANOVA example

Lecture 165 Code: One-way ANOVA (independent samples)

Lecture 166 Code: One-way repeated-measures ANOVA

Lecture 167 Two-way ANOVA example

Lecture 168 Code: Two-way mixed ANOVA

Section 14: Regression

Lecture 169 Introduction to GLM / regression

Lecture 170 Least-squares solution to the GLM

Lecture 171 Evaluating regression models: R2 and F

Lecture 172 Simple regression

Lecture 173 Code: simple regression

Lecture 174 "Unsupervised learning": Compute R2 and F

Lecture 175 Multiple regression

Lecture 176 Standardizing regression coefficients

Lecture 177 Code: Multiple regression

Lecture 178 Polynomial regression models

Lecture 179 Code: polynomial modeling

Lecture 180 "Unsupervised learning": Polynomial design matrix

Lecture 181 Logistic regression

Lecture 182 Code: Logistic regression

Lecture 183 Under- and over-fitting

Lecture 184 "Unsupervised learning": Overfit data

Lecture 185 Comparing "nested" models

Lecture 186 What to do about missing data

Section 15: Statistical power and sample sizes

Lecture 187 What is statistical power and why is it important?

Lecture 188 Estimating statistical power and sample size

Lecture 189 Compute power and sample size using G*Power

Section 16: Clustering and dimension-reduction

Lecture 190 K-means clustering

Lecture 191 Code: k-means clustering

Lecture 192 "Unsupervised learning:" K-means and normalization

Lecture 193 "Unsupervised learning:" K-means on a Gauss blur

Lecture 194 Clustering via dbscan

Lecture 195 Code: dbscan

Lecture 196 "Unsupervised learning": dbscan vs. k-means

Lecture 197 K-nearest neighbor classification

Lecture 198 Code: KNN

Lecture 199 Principal components analysis (PCA)

Lecture 200 Code: PCA

Lecture 201 "Unsupervised learning:" K-means on PC data

Lecture 202 Independent components analysis (ICA)

Lecture 203 Code: ICA

Section 17: Signal detection theory

Lecture 204 The two perspectives of the world

Lecture 205 d-prime

Lecture 206 Code: d-prime

Lecture 207 Response bias

Lecture 208 Code: Response bias

Lecture 209 F-score

Lecture 210 Receiver operating characteristics (ROC)

Lecture 211 Code: ROC curves

Lecture 212 "Unsupervised learning": Make this plot look nicer!

Section 18: A real-world data journey

Lecture 213 Note about the code for this section

Lecture 214 Introduction

Lecture 215 MATLAB: Import and clean the marriage data

Lecture 216 MATLAB: Import the divorce data

Lecture 217 MATLAB: More data visualizations

Lecture 218 MATLAB: Inferential statistics

Lecture 219 Python: Import and clean the marriage data

Lecture 220 Python: Import the divorce data

Lecture 221 Python: Inferential statistics

Lecture 222 Take-home messages

Section 19: Bonus section

Lecture 223 About deep learning

Lecture 224 Bonus content