Explore practical coding of basic machine learning models using Python. Gain insights into algorithms like linear regression, logistic regression, SVM, KNN, and decision trees.

Machine Learning with Python A Practical Beginners’ Guide by Oliver Theobald-01.png

MLtarball.tar.gz

jupyter1

jupyter2

jupyter3

jupyter4

jupyter5

jupyter6

jupyter7

jupyter8

jupyter9

jupyter10

jupyter11

jupyter12

jupyter13

jupyter14

jupyter15

jupyter16

jupyter17

jupyter18dtextra

jupyter19dt

jupyter20rf

jupyter21gbc

jupyter22gbr

jupyter-temp

pca6

split

lor1

lor2

lor3

lor4

lor5

lor6

lor7

lor8

lor9

svm9

knn9

gbc7

gbr1

gbr2

gbr3

gbr4

gbr5

gbr6

gbr7

gbr8

Mlxtend

This course teaches you how to code basic machine learning models. The content is designed for beginners with general knowledge of machine learning, including common algorithms such as linear regression, logistic regression, SVM, KNN, decision trees, and more. If you need a refresher, we have summarized key concepts from machine learning, and there are overviews of specific algorithms dispersed throughout the course.

A Practical Guide to Machine Learning with Python

Support vector machines are considered some of the best classifiers in supervised learning for analyzing complex data and downplaying the influence of outliers.

Developed within the computer science community in the 1990s, SVM was originally designed for predicting numeric and categorical outcomes as a double-barrel prediction technique. Today, SVM is mostly used as a classification technique for predicting categorical outcomes, similar to logistic regression.

SVM mirrors logistic regression in binary prediction scenarios as it attempts to separate classes based on the mathematical relationship between variables. However, unlike logistic regression, SVM attempts to separate data classes by maximum distance between the partitioned data points.

> Its key feature is the margin, the distance between the boundary line and the nearest data point, multiplied by two. The margin is able to cope with new data points and outliers that would infringe on a logistic regression boundary line.

# Example
Given the following positively labelled data points {(1,1), (2,1), (1,-1), (2,-1)}
and the following negative labelled data points {(4,0), (5,1), (5,-1), (6,0)}


- Plot all given data points.



- Once you plot the data points, you need to identify the nearest data points. Two data points, (2,1) and (2,-1), are negatively labeled data points and one, (4,0),  is a positively labeled data point. Based on this, we know these are the closest to each other.

$s1=\binom{2}{1}$, $s2=\binom{2}{-1}$ and $s3=\binom{4}{0}$

Now write the equation of the hyperplane. Each vector is augmented with a 1 as the bias input. 
$s1'=\binom{\binom{2}{1}}{1}$, $s2'=\binom{\binom{2}{-1}}{1}$ and $s3'=\binom{\binom{4}{0}}{1}$

$\alpha_1s_1'.s1' + \alpha_2s_2'.s1' + \alpha_3s_3'.s1'=-1$

$\alpha_1s_1'.s2' + \alpha_2s_2'.s2' + \alpha_3s_3'.s3'=-1$

$\alpha_1s_1'.s3' + \alpha_2s_2'.s3' + \alpha_3s_3'.s3'=1$

- Now calculate the values of $\alpha_1, \alpha_2$  and $\alpha_3$.

$\alpha_1 \binom{\binom{2}{1}}{1} . \binom{\binom{2}{1}}{1} + \alpha_2 \binom{\binom{2}{-1}}{1} . \binom{\binom{2}{1}}{1} + \alpha_3 \binom{\binom{4}{0}}{1} . \binom{\binom{2}{1}}{1}=-1$

$\alpha_1 \binom{\binom{2}{1}}{1} . \binom{\binom{2}{-1}}{1} + \alpha_2 \binom{\binom{2}{-1}}{1} . \binom{\binom{2}{-1}}{1} + \alpha_3 \binom{\binom{4}{0}}{1} . \binom{\binom{2}{-1}}{1}=-1$

$\alpha_1 \binom{\binom{2}{1}}{1} . \binom{\binom{4}{0}}{1} + \alpha_2 \binom{\binom{2}{-1}}{1} . \binom{\binom{4}{0}}{1} + \alpha_3 \binom{\binom{4}{0}}{1} . \binom{\binom{4}{0}}{1}=-1$

$6\alpha_1 + 4\alpha_2 + 9\alpha_3=-1$

$4\alpha_1 + 6\alpha_2 + 9\alpha_3=-1$

$9\alpha_1 + 9\alpha_2 + 17\alpha_3=1$

After solving the three equations:
$\alpha_1=-3.25, \alpha_2=-3.25$ and $\alpha_3=3.5$

**Now calculate the weight vector.**
<div align="center">

$w'=\sum_{i=1}^{3} \alpha_is_i$

</div>

$w'=-3.25\binom{\binom{2}{1}}{1} + (-3.25)\binom{\binom{2}{-1}}{1} + 3.5\binom{\binom{4}{0}}{1}$

$w'=\binom{\binom{1}{0}}{-3}$

**Hyperplane equation**
<div align="center">

$y=wx+b$

</div>

$w=\binom{1}{0}$ (parallel to y-axis)

and 

$b=-3$ or $b+3=0$






This lesson will provide an overview of the support vector machines and the steps involved in its implementation.

Implementation of Support Vector Machines

Introduction to Course

Introduction to Machine Learning

Exploratory Data Analysis

Data Scrubbing

Pre-Model Algorithms

Split Validation

Model Design

Linear Regression

Logistic Regression

Support Vector Machines

K-Nearest Neighbors

Tree-Based Methods

Conclusion

Appendix

Implementation of Support Vector Machines

Example