Machine Learning For Aspiring Data Scientists: Zero To Hero

Machine Learning For Aspiring Data Scientists: Zero To Hero - Druckversion

+- Forum Rockoldies (https://rockoldies.net/forum)
+-- Forum: Fotobearbeitung - Photoshop (https://rockoldies.net/forum/forumdisplay.php?fid=16)
+--- Forum: E-Learning, Tutorials (https://rockoldies.net/forum/forumdisplay.php?fid=18)
+--- Thema: Machine Learning For Aspiring Data Scientists: Zero To Hero (/showthread.php?tid=64828)

Machine Learning For Aspiring Data Scientists: Zero To Hero - Panter - 17.12.2022

[Bild: sxjpw3xd5kypp1twqnorvh9iow.jpg]

Machine Learning For Aspiring Data Scientists: Zero To Hero
Published 8/2022
MP4 | Video: h264, 1280x720 | Audio: AAC, 44.1 KHz
Language: English | Size: 9.23 GB | Duration: 16h 7m

Learn the foundations of machine learning necessary to get a job in data science. No coding experience required.

What you'll learn
Undertand the foundations of machine learning even if you're a total beginner
Be able to pass job interviews for data science jobs
Learn without wasting time in things that don't come up in interviews or real work
Avoid rookie mistakes that waste companies' time and money

Requirements
No programming or advanced math experience required! You'll learn everything you need to know.

Description
This course will teach you the complete foundations of machine learning that you need to get a job in data science (and do a great job afterward). The course will help you:Pass job interviews and technical quizzesAvoid rookie mistakes that waste companies' time and moneyBe prepared for real work.Important stuff about this course:You won't spend hours learning stuff that never comes up in a job interview.Total beginners are welcome; coding experience or advanced math knowledge are not required.It was designed by an industry expert who's been on the hiring side of the table and knows what companies are looking for.This course will be of great help if you are:A student who wants to prepare for work in data science after graduating.An established professional or academic who wants to switch careers to data science.A total beginner who wants to dabble in machine learning and data science for the first time.How is this different from an academic course or a bootcamp?In academic courses, your teacher spends hours speaking about calculus and linear algebra, but then none of that comes up in a job interview! That in-depth knowledge certainly has a place but is not what most companies are looking for.In bootcamps you tend to learn how to use many tools but not how they work under the hood. This black-box knowledge is what companies want to avoid the most in applicants!This course sits in between-you gain foundational knowledge and truly understand machine learning, without spending time on unimportant stuff.
Overview
Section 1: Machine Learning Models
Lecture 1 Modeling an epidemic
Lecture 2 The machine learning recipe
Lecture 3 The components of a machine learning model
Lecture 4 Why model?
Lecture 5 On assumptions and can we get rid of them?
Lecture 6 The case of AlphaZero
Lecture 7 Overfitting/underfitting/bias/variance
Lecture 8 Why use machine learning
Lecture 9 Notes on machine learning models
Section 2: Linear regression
Lecture 10 The InsureMe challenge
Lecture 11 Supervised learning
Lecture 12 A quick note on the word "features"
Lecture 13 Linear assumption
Lecture 14 Linear regression template
Lecture 15 Non-linear vs proportional vs linear
Lecture 16 Linear regression template revisited
Lecture 17 Loss function
Lecture 18 Training algorithm
Lecture 19 Code time
Lecture 20 R squared
Lecture 21 Why use a linear model?
Lecture 22 Kaggle notebook on linear regression
Lecture 23 Notes on supervised learning and linear regression
Lecture 24 Finding closed-form solution to linear regression (optional)
Section 3: Scaling and Pipelines
Lecture 25 Introduction to scaling
Lecture 26 Min-max scaling
Lecture 27 Code time (min-max scaling)
Lecture 28 The problem with min-max scaling
Lecture 29 What's your IQ?
Lecture 30 Standard scaling
Lecture 31 Code time (standard scaling)
Lecture 32 Model before and after scaling
Lecture 33 Inference time
Lecture 34 Pipelines
Lecture 35 Code time (pipelines)
Lecture 36 Kaggle notebook on scaling and pipelines
Lecture 37 Notes on scaling and pipelines
Section 4: Regularization
Lecture 38 Spurious correlations
Lecture 39 L2 regularization
Lecture 40 Code time (L2 regularization)
Lecture 41 L2 results
Lecture 42 L1 regularization
Lecture 43 Code time (L1 regularization)
Lecture 44 L1 results
Lecture 45 Why does L1 encourage zeros?
Lecture 46 L1 vs L2: Which one is best?
Lecture 47 Kaggle notebook on regularization
Lecture 48 Notes on regularization
Section 5: Validation
Lecture 49 Introduction to validation
Lecture 50 Why not evaluate model on training data
Lecture 51 The validation set
Lecture 52 Code time (validation set)
Lecture 53 Error curves
Lecture 54 Model selection
Lecture 55 The problem with model selection
Lecture 56 Tainted validation set
Lecture 57 Monkeys with typewriters
Lecture 58 My own validation epic fail
Lecture 59 The test set
Lecture 60 What if the model doesn't pass the test?
Lecture 61 How not to be fooled by randomness
Lecture 62 Cross-validation
Lecture 63 Code time (cross validation)
Lecture 64 Cross-validation results summary
Lecture 65 AutoML
Lecture 66 Is AutoML a good idea?
Lecture 67 Red flags: Don't do this!
Lecture 68 Red flags summary and what to do instead
Lecture 69 Your job as a data scientist
Lecture 70 Kaggle notebook on validation and cross-validation
Lecture 71 30-minute code assignment with new dataset!
Lecture 72 Notes on validation and testing
Lecture 73 Extra reading: Model retraining
Lecture 74 Extra reading: The Difference between Statistics and Machine Learning
Section 6: Common Mistakes
Lecture 75 Intro and recap
Lecture 76 Mistake #1: Data leakage
Lecture 77 The golden rule
Lecture 78 Helpful trick (feature importance)
Lecture 79 Real example of data leakage (part 1)
Lecture 80 Real example of data leakage (part 2)
Lecture 81 Another (funny) example of data leakage
Lecture 82 Mistake #2: Random split of dependent data
Lecture 83 Another example (insurance data)
Lecture 84 Mistake #3: Look-Ahead Bias
Lecture 85 Example solutions to Look-Ahead Bias
Lecture 86 Consequences of Look-Ahead Bias
Lecture 87 How to split data to avoid Look-Ahead Bias
Lecture 88 Cross-validation with temporally related data
Lecture 89 Mistake #4: Building model for one thing, using it for something else
Lecture 90 Sketchy rationale
Lecture 91 Why this matters for your career and job search
Lecture 92 Find the error: 10-minute code assignment
Lecture 93 Assignment solution
Lecture 94 Notes on common mistakes
Section 7: Classification - Part 1: Logistic Model
Lecture 95 Classifying images of handwritten digits
Lecture 96 Why the usual regression doesn't work
Lecture 97 Machine learning recipe recap
Lecture 98 Logistic model template (binary)
Lecture 99 Decision function and boundary (binary)
Lecture 100 Logistic model template (multiclass)
Lecture 101 Decision function and boundary (multi-class)
Lecture 102 Summary: binary vs multiclass
Lecture 103 Code time!
Lecture 104 Why the logistic model is often called logistic regression
Lecture 105 One vs Rest, One vs One
Lecture 106 Kaggle notebook on logistic model for digit classification
Lecture 107 Notes on Logistic Model
Section 8: Classification - Part 2: Maximum Likelihood Estimation
Lecture 108 Where we're at
Lecture 109 Brier score and why it doesn't work
Lecture 110 The likelihood function
Lecture 111 Optimization task and numerical stability
Lecture 112 Let's improve the loss function
Lecture 113 Loss value examples
Lecture 114 Adding regularization
Lecture 115 Binary cross-entropy loss
Lecture 116 Notes on Maximum Likelihood Estimation
Section 9: Classification - Part 3: Gradient Descent
Lecture 117 Recap
Lecture 118 No closed-form solution
Lecture 119 Naive algorithm
Lecture 120 Fog analogy
Lecture 121 Gradient descent overview
Lecture 122 The gradient
Lecture 123 Numerical calculation
Lecture 124 Parameter update
Lecture 125 Convergence
Lecture 126 Analytical solution
Lecture 127[Optional] Interpreting analytical solution
Lecture 128 Gradient descent conditions
Lecture 129 Beyond vanilla gradient descent
Lecture 130 Code time
Lecture 131 Reading the documentation
Lecture 132 10-minute coding exercise: Classify images of clothes
Lecture 133 Notes on Gradient Descent
Section 10: Classification metrics and class imbalance
Lecture 134 Binary classification and class imbalance
Lecture 135 Assessing performance
Lecture 136 Accuracy
Lecture 137 Accuracy with different class importance
Lecture 138 Precision and Recall
Lecture 139 Sensitivity and Specificity
Lecture 140 F-measure and other combined metrics
Lecture 141 ROC curve
Lecture 142 Area under the ROC curve
Lecture 143 Custom metric (important stuff!)
Lecture 144 Other custom metrics
Lecture 145 Bad data science process :(
Lecture 146 Data rebalancing (avoid doing this!)
Lecture 147 Stratified split
Lecture 148 Notes on Classification Metrics
Section 11: Neural Networks
Lecture 149 The inverted MNIST dataset
Lecture 150 The problem with linear models
Lecture 151 Neurons
Lecture 152 Multi-layer perceptron (MLP) for binary classification
Lecture 153 MLP for regression
Lecture 154 MLP for multi-class classification
Lecture 155 Hidden layers
Lecture 156 Activation functions
Lecture 157 Decision boundary
Lecture 158 Loss function
Lecture 159 Intro to neural network training
Lecture 160 Parameter initialization
Lecture 161 Saturation
Lecture 162 Non-convexity
Lecture 163 Stochastic gradient descent (SGD)
Lecture 164 More on SGD
Lecture 165 Code time!
Lecture 166 Backpropagation
Lecture 167 The problem with MLPs
Lecture 168 Deep learning
Lecture 169 Notes on Neural Networks
Lecture 170 20-minute coding task
Section 12: Tree-Based Models
Lecture 171 Decision trees
Lecture 172 Building decision trees
Lecture 173 Stopping tree growth
Lecture 174 Pros and cons of decision trees
Lecture 175 Decision trees for classification
Lecture 176 Decision boundary
Lecture 177 Bagging
Lecture 178 Random forests
Lecture 179 Gradient-boosted trees for regression
Lecture 180 Gradient-boosted trees for classification[optional]
Lecture 181 How to use gradient-boosted trees
Lecture 182 20-minute coding exercise (important!)
Section 13: K-nn and SVM
Lecture 183 Nearest neighbor classification
Lecture 184 K nearest neighbors
Lecture 185 Disadvantages of k-NN
Lecture 186 Recommendation systems (collaborative filtering)
Lecture 187 Introduction to Support Vector Machines (SVMs)
Lecture 188 Maximum margin
Lecture 189 Soft margin
Lecture 190 SVM vs Logistic Model (support vectors)
Lecture 191 Alternative SVM formulation
Lecture 192 Dot product
Lecture 193 Non-linearly separable data
Lecture 194 Kernel trick (polynomial)
Lecture 195 RBF kernel
Lecture 196 SVM remarks
Section 14: Unsupervised Learning
Lecture 197 Intro to unsupervised learning
Lecture 198 Clustering
Lecture 199 K-means clustering
Lecture 200 K-means application example
Lecture 201 Elbow method
Lecture 202 Clustering remarks
Lecture 203 Intro to dimensionality reduction
Lecture 204 PCA (principal component analysis)
Lecture 205 PCA remarks
Lecture 206 Code time (PCA)
Section 15: Feature Engineering
Lecture 207 Missing data
Lecture 208 Imputation
Lecture 209 Imputer within pipeline
Lecture 210 One-Hot encoding
Lecture 211 Ordinal encoding
Lecture 212 How to combine pipelines
Lecture 213 Code sample
Lecture 214 Feature Engineering
Lecture 215 Features for Natural Language Processing (NLP)
Lecture 216 Anatomy of a Data Science Project
Lecture 217 Next steps!
Lecture 218 Final Project: Predict Titanic survivors
Aspiring data scientists who want to get their first job in the field.,Software engineers who want to be involved in data science and machine learning.,Researchers who want to make the move from academia to industry

Show Content