Introduction
Figure 1
An infographic showing the relationships between
AI, ML, and DL
Figure 2
Types of Machine Learning
Figure 3
Summary of key concepts and sub-domains of
ML
Supervised methods - Regression
Figure 1
Example of linear and polynomial
regressions
Figure 2
Inspection of our dataset
Figure 3
Linear regression of dataset I
Figure 4
Linear regression of dataset II
Figure 5
Linear regression of dataset III
Figure 6
Linear regression of dataset IV
Figure 7
Comparison of the regressions of our
dataset
Figure 8
Polynomial regression of dataset I
Figure 9
Polynomial regression of dataset II
Figure 10
Polynomial regression of dataset III
Figure 11
Polynomial regression of dataset IV
Figure 12
Polynomial regression of dataset I with N
between 1 and 10
Figure 13
Polynomial regression of dataset II with N
between 1 and 10
Figure 14
Polynomial regression of dataset III with N
between 1 and 10
Figure 15
Polynomial regression of dataset IV with N
between 1 and 10
Figure 16
Comparison of the regressions of our
dataset
Figure 17
Comparison of the regressions of our
dataset
Supervised methods - classification
Figure 1
Artwork by @allison_horst
Figure 2
Artwork by @allison_horst
Figure 3
Visualising the penguins dataset
Figure 4
Visualising the penguins dataset
Figure 5
Decision tree for classifying penguins
Figure 6
Decision tree for classifying penguins
Figure 7
Classification space for our decision tree
Figure 8
Performance of decision trees of various
depths
Figure 9
Simplified decision tree
Figure 10
Classification space of the simplified decision
tree
Figure 11
Classification space generated by the SVM
model
Ensemble Methods
Figure 1
Stacking
Figure 2
Bagging
Figure 3
Boosting
Figure 4
Random Forests
Figure 5
Random forest trees
Figure 6
Random forest clf space
Figure 7
Regressor predictions and average from
stack
Unsupervised methods - Clustering
Figure 1
Plot of the random clusters
Figure 2
Plot of the fitted random clusters
Figure 3
An example of kmeans failing on non-linear
cluster boundaries
Figure 4
Kmeans attempting to classify overlapping
clusters
Figure 5
Spectral clustering on two concentric
circles
Figure 6
Spectral clustering viewed with an extra
dimension
Figure 7
Kmeans attempting to cluster the concentric
circles
Figure 8
Spectral clustering on the concentric
circles
Unsupervised methods - Dimensionality reduction
Figure 1
MNIST example illustrating all the classes in
the dataset
Figure 2
MNIST example of a single image
Figure 3
SKLearn image with highlighted pixels
Figure 4
SKLearn image with highlighted pixels
Figure 5
Reduction using PCA
Figure 6
Reduction using PCA, with K-means
clustering
Figure 7
Reduction using PCA, adding colour
labelling
Figure 8
Reduction using PCA, applying t-SNE to the MNIST
data
Figure 9
Reduction using PCA, running K-means clustering
on new 2D representation
Figure 10
Reduction using PCA, adding in colour
labelling
Figure 11
Reduction to 3 components using pca
Figure 12
Reduction to 3 components using tsne
Neural Networks
Figure 1
A diagram of a perceptron
Figure 2
A multi-layer perceptron