Applied Machine Learning in R
- Description
- Curriculum
- FAQ
- Reviews
This course offers you practical training in machine learning, using the R program. At the end of the course you will know how to use the most widespread machine learning techniques to make accurate predictions and get valuable insights from your data.
All the machine learning procedures are explained live, in detail, on real life data sets. So you will advance fast and be able to apply your knowledge immediately – no need for painful trial-and-error to figure out how to implement this or that technique in R. Within a short time you can have a solid expertise in machine learning.
Machine learning skills are very valuable if you intent to secure a job like data analyst, data scientist, researcher or even software engineer. So it may be the right time for you to enroll in this course and start building your machine learning competences today!
Let’s see what you are going to learn here.
First of all, we are going to discuss some essential concepts that you must absolutely know before performing machine learning. So we’ll talk about supervised and unsupervised machine learning techniques, about the distinctions between prediction and inference, about the regression and classification models and, above all, about the bias-variance trade-off, a crucial issue in machine learning.
Next we’ll learn about cross-validation. This is an all-important topic, because in machine learning we must be able to test and validate our model on independent data sets (also called first seen data). So we are going to present the advantages and disadvantages of three cross-validations approaches.
After the first two introductory sections, we will get to study the supervised machine learning techniques. We’ll start with the regression techniques, where the response variable is quantitative. And no, we are not going to stick to the classical OLS regression that you probably know already. We will study sophisticated regression techniques like stepwise regression (forward and backward), penalized regression (ridge and lasso) and partial least squares regression. And of course, we’ll demonstrate all of them in R, using actual data sets.
Afterwards we’ll go to the classification techniques, very useful when we have to predict a categorical variable. Here we’ll study the logistic regression (classical and lasso), discriminant analysis (linear and quadratic), naïve Bayes technique, K nearest neighbor, support vector machine, decision trees and neural networks.
For each technique above, the presentation is structured as follows:
* a short, easy to understand theoretical introduction (without complex mathematics)
* how to train the predictive model in R
* how to test the model to make sure that it does a good prediction job on independent data sets.
In the last sections we’ll study two unsupervised machine learning techniques: principal component analysis and cluster analysis. They are powerful data mining techniques that allow you to detect patterns in your data or variables.
For each technique, a number of practical exercises are proposed. By doing these exercises you’ll actually apply in practice what you have learned.
This course is your opportunity to become a machine learning expert in a few weeks only! With my video lectures, you will find it very easy to master the major machine learning techniques. Everything is shown live, step by step, so you can replicate any procedure at any time you need it.
So click the “Enroll” button to get instant access to your machine learning course. It will surely provide you with new priceless skills. And, who knows, it could give you a tremendous career boost in the near future.
See you inside!
-
2
What Is Machine Learning?A definition of machine learning and a description of the machine learning process.
-
3
Supervised vs. Unsupervised Methods
Maybe the most important model categorization of machine learning models: models that have a target variable (supervised) vs. models that do not have one (unsupervised).
-
4
Prediction vs. Inference
There are two purposes we can use machine learning models: prediction and inference. Here we'll discuss the difference between them.
-
5
Restrictive Models vs. Flexible Models
Should we prefer a simple restrictive ML model or a more complex one? Here we discuss the pros and cons for each.
-
6
Computing Prediction Accuracy of Regression Models
Here we present a couple of indicators that measure the prediction performance for the regression models.
-
7
Computing Prediction Accuracy of Classification Models
The best indicators that help us estimate the predictive performance of the classification models.
-
8
Bias-Variance Tradeoff
This is a crucial issue in machine learning. Some models may have low bias, but high variance, other models may present high bias, but lower variance. Which is best? How to select an "optimal" model?
-
9
What Is Cross-Validation?
Machine learning models should work well on independent data sets. This is why it is absolutely necessary to validate them before actually using them.
-
10
Validation Set Approach
The advantages and disadvantages of the validation set approach.
-
11
Leave-One-Out Cross-Validation Approach
What is leave-one-out cross validation and what are its strengths and weaknesses.
-
12
K-Fold Cross-Validation Approach
What is k-fold cross-validation, how and when we should use it.
-
13
Introduction to the OLS Regression
How to run an ordinary least squares regression in R, in order to predict the values of a numeric (quantitative) variable.
-
14
Validating the OLS Regression Model (1)
How to validate an OLS regression model using the validation set approach.
-
15
Validating the OLS Regression Model (2)
How to validate an OLS regression model using the k-fold cross-validation method.
-
16
Best Subset Selection Regression - Introduction
When we have a lot of predictors in our regression model, we are often interested to identify the most relevant of them. Here we explain how we can do that using the best subset regression.
-
17
Forward Selection Regression
How to perform a forward selection regression in order to find the predictors that most influence the response variable.
-
18
Backward Selection Regression
How to perform a backward forward selection regression, with the same goal: find the predictors that have the strongest effect on the response variable.
-
19
Validating the Subset Selection Regression
How to validate the subset selection regression on an independent data set.
-
20
Ridge Regression
How to run a ridge regression in R and interpret the output.
-
21
Validating the Ridge Regression
How to validate the ridge regression using the k-fold cross-validation approach.
-
22
Lasso Regression
How to perform a lasso regression in R and interpret the results.
-
23
Validating the Lasso Regression
How to validate the lasso regression model with the k-fold cross-validation method.
-
26
Introduction to Logistic Regression
How to run the logistic regression in R and interpret the coefficients.
-
27
Computing the Prediction Accuracy
How to compute the percentage of correctly classified cases for a logistic regression (in the whole sample).
-
28
Building the ROC Curve
How to draw and interpret the ROC curve - one of the most important output of a logistic regression.
-
29
Validating the Logistic Regression
How to test a logistic regression model in an independent data set.
-
30
Lasso Logistic Regression
If our logistic regression model has many predictors, we can use the lasso logistic regression to select the most effective of them. In this lecture we learn how to do that.
-
31
Validating the Lasso Logistic Regression
How to validate a lasso logistic regression model - and select the model with the highest prediction accuracy.
-
32
Linear Discriminant Analysis
How to run a linear discriminant analysis and interpret the output.
-
33
Validating the Linear Discriminant
How to test (validate) our linear discriminant analysis.
-
34
Quadratic Discriminant Analysis
How to execute a quadratic discriminant analysis in R.
-
35
Validating the Quadratic Discriminant
How to validate our quadratic discriminant model (check if it does a good prediction job in independent data sets).
-
36
Introduction to Naive Bayes Estimation
Here we explain in detail how the naive Bayes procedure works, and give a simple illustration in a hypothetical situation.
-
37
Naive Bayes Estimation in R with the e1071 Package
How to apply the naive Bayes classification in R using the e1071 package.
-
38
Validating the Naive Bayes Model
How to validate a naive Bayes estimation model.
-
39
Naive Bayes Estimation in R with the naivebayes Package
How to perform a naive Bayes classification in R using the naivebayes package (an how to test our model).
-
40
Introduction to K Nearest Neighbor
What are the principles of the K-nearest neighbor technique and how it works in practice.
-
41
K Nearest Neighbor in R
How to actually perform the K-nearest neighbor predictive analysis in R, and how to compute the predictive accuracy in the test set.
-
42
Finding the Optimal Number of Neighbors
The result of the KNN analysis depends on the number of neighbors we consider (K). In this lecture we learn how to find the number of neighbors that ensures the greatest predictive accuracy in the validation set.
-
43
Maximal Margin Classifier
The maximal margin classifier is the simplest version of a support vector machine. In this lecture we'll learn when we can use it.
-
44
Support Vector Classifier
A more applicable version of SVM is the support vector classifier. Let's see what it is and when we can use this technique.
-
45
Introduction to Support Vector Machine
What is support vector machine, what is the "kernel trick" and how many types of kernels exist in SVM. Everything with illustrations.
-
46
Support Vector Machine with Linear Kernel
Everything about the SVM model with a linear boundary between classes: how to implement it in R, how to validate it, how to find the optimal cost value.
-
47
Support Vector Machine with Polynomial Kernel
How to perform an SVM analysis where the boundary between classes is polynomial, how to test our model and how to find the optimal polynomial degree.
-
48
Support Vector Machine with Radial Kernel
How to run an SVM analysis when the boundary between classes is closed (radial), how to test our model and how to find the optimal value for the gamma parameter.
-
49
What Are Decision Trees?
A definition of decision trees and a detailed explanation on how they work.
-
50
Introduction to CART
What are regression tress and classification trees, and how to compute the predictive accuracy for them.
-
51
Advantages and Disadvantages of Decision Trees
Decision trees do have advantages, but they also have a couple of important weaknesses. We discuss them all in this lecture.
-
52
Growing Regression Trees in R
How to build a regression tree in R, and how to validate it.
-
53
Growing Classification Trees in R
How to build a classification tree in R, and how to validate it.
-
54
Introduction to Pruning
Sometimes we can improve the prediction accuracy of a tree by simplifying it, i.e. by cutting some branches. In this lecture we show how we could do that.
-
55
Pruning Regression Trees in R
Practical pruning example on the regression tree we have previously created.
-
56
Pruning Classification Trees in R
Practical pruning example on the classification tree we have previously created.
-
57
Introduction to Bagging and Random Forests
Other ways to improve the prediction accuracy of decision trees is to create many trees and aggregate their outputs. Bagging and random forests are two techniques that perform this task.
-
58
Bagging Regression Trees in R
Use bagging with the regression tree we already created, and interpret the results.
-
59
Bagging Classification Trees in R
Use bagging with the classification tree we already created, and interpret the results.
-
60
Random Forests of Regression Trees in R
Grow a random forest of regression trees and compute its prediction accuracy in the test set.
-
61
Random Forests of Classification Trees in R
Grow a random forest of classification trees and compute its prediction accuracy in the test set.
-
62
Introduction to Boosting
Another technique meant to increase the predictive performance of decision trees.
-
63
Boosting Regression Trees
Use the boosting method with the regression tree we already created.
-
64
Boosting Classification Trees
Use the boosting method with the classification tree we already created.
-
65
Definition of Neural Networks
What are neural networks and what is a multilayer perceptron - the most popular type of network.
-
66
What Happens Inside of a Neuron?
Here we'll learn how a neural network works: what operations are performed in a hidden layer neuron.
-
67
Neural Network Learning Process
How exactly a neural network learns how to predict our response variable, step by step.
-
68
A Simple Neural Network Example
In this lecture we give an example of neural network using simulated data, so you understand how to build and train a neural network in R.
-
69
Practical Neural Network Example
Create and train a neural network that is actually used to predict whether a phone company customer would abandon the company in the future or not.
-
70
Practical Neural Network Example (2)
Here we modify some parameters of the neural network we have previously created, to see whether we can get a higher predictive performance in the test set.
-
71
Introduction to Principal Component Analysis
What is principal component analysis and when it is used.
-
72
Executing the Principal Component Analysis
How to perform PCA in R (on a practical, real example) and how to identify and interpret the components.
-
73
Performing the Adequacy Tests
How to check whether our PCA model is strong enough (i.e. whether our components are relevant for the initial variables).
-
74
Introduction to Cluster Analysis
What is cluster analysis and what are the differences between hierarchical cluster and k-means cluster.
-
75
Hierarchical Cluster
How to run a hierarchical cluster in R, identify the clusters and describe them.
-
76
K-Means Cluster
How to perform a k-means cluster in R and create a profile for each cluster.
