Machine Learning For Data Science

Core Concepts Overview Preview 01:29

What will we discover with core concepts?  Here I'll give you a brief overview of all the exciting lectures contained in this section.

Computer Science - the `Train Wreck' Definition Preview 00:54

The current definition of computer science is an incomprehensible train wreck!  Find out why in this lecture! 

What's Data / "I can see data everywhere!" Preview 05:34

In order to better understand what computer science is, it's useful to understand what DATA is.  By the end of this lecture you'll be able to see DATA EVERYWHERE you look! 

Structured vs Unstructured Data Preview 02:44

There are two different kinds of data - Structured and Unstructured.  This is a key concept, that we are going to come back to time and time again later on.  Important, and delivered in under 3 minutes! 

Structured and Unstructured Data

Test your understanding of structured vs. unstructured data in this quick quiz!

Computer Science - Definition Revisited & The Greatest "lie" ever SOLD.... Preview 11:03

Here we revisit the definition of what Computer Science is, with something that's actually comprehensible.  Wondering what an algorithm is?  We've got that covered to?  And while we're at it - we'll even dive into programming.

Finally, we'll touch on what I call "One of the greatest lies - Ever SOLD".

What's big data? Preview 06:39

So what is Big Data?  Learn the three V's of big data, what it is... and what it isn't!  

This lecture will educate you so you don't fall for the "marketing hype" often associated with Big Data. 

Big Data - Quiz

A quiz on the ideas of big data.

What is Artificial Intelligence (AI) Preview 11:57

This is a longer lecture, however within 12 minutes we'll cover off the most fundamental parts of Artificial Intelligence. 

Do you how a computer plans a route in a GPS?  Or how it would play a game like Tic-Tac-Toe?  The answers might surprise you!  This lecture has several animations to help illustrate the concepts and importantly - the challenges of AI in search. 

And!  We'll also cover off one of the most interesting questions - "Can a computer Really Think"?

What is Machine Learning? - Part 1 - The ideas Preview 11:07

Alas!  We are discussing Machine Learning!  In this lecture, we'll clearly define Machine Learning.  We'll give a simplified overview of the Machine Learning Process, which we'll expand later on in section 4.  We'll discuss some applications of Machine Learning, as well as what Machine Learning gives AI.

By the end of this lecture, you'll have an idea of what Machine Learning can be used for.

What is Machine Learning? - Part 2 - An Example Preview 06:48

In this Animated Example, we'll show a simple Machine Learning application.  While it's a very simple example, it will show how data can be looked at, examined for patterns, and will discuss the difference between sensitivity and specificity.  These are key concepts to Machine Learning and important to understand when applying it.

What is data science? Preview 05:05

What is Data Science?  Magical Unicorns?  (Yes really!).  Battling Venn Diagrams (I'm not kidding!)

In this lecture, we'll define what Data Science is and what a Data Scientist does. 

Recap & How do these relate to each other? Preview 03:27

Big Data!  AI!  Machine Learning!  Computer Science!  Data Science!  

How does this all fit together?  Where does one "start" and the other "stop?"  In this lecture, we'll use an animated diagram to explain how all these different domains interrelate.  Confusion stops here! 

Impacts, Importance and examples - Overview Preview 00:44

What will we discover with "Impacts, Importance and Examples"?  Here I'll give you a brief overview of all the exciting lectures contained in this section.

Why is this important now? Preview 07:15

Why are we talking about this?  Why is this important now! 

In this lecture we'll uncover the convergence of events that have come together in a perfect storm of digital change. 

Computers exploding! - The explosive growth of computer power explained. Preview 13:23

Computers exploding?!  Every one always gives lip service to "how much technology has actually changed".   But what does it really mean?  In this longer lecture, we'll take a journey from 1956 to 2014, and really explain how the world has changed.

Do you have a super computer in your house?  You might be surprised to find out the truth.....!

What problems does Machine Learning Solve? Preview 04:29

In this brief lecture, we'll cover the three different problems Machine Learning solves really well. 

• Classification
• Clustering
• Regression

Pictures will help make sense of every concept, and it will be the bedrock for later seeing how different problems can be solved by Machine Learning.  While watching this lecture, be sure to look at how a problem can be solved in different ways, using different approaches to Machine Learning.

Where it's transforming our lives Preview 09:16

We've covered off - what it is.  How it works.  What it provides....

Now the question is How is this changing our lives?

In this lecture we'll talk about what we'll likely see.  What happens when Machine Learning goes wrong.  And we'll touch on ethics - which is not just a case of banning killer robots, but much more subtle as well.

The Machine Learning Process - Overview Preview 01:14

What will we discover with "The Machine Learning Process"?  Here I'll give you a brief overview of all the exciting lectures contained in this section.

5 Step Machine Learning Process Overview Preview 01:53

In this lecture we'll cover off each of the five step of the Machine Learning Process, sometimes called a "pipeline" or "workflow".  Any problem being solved by Machine Learning will have to touch all of these fives steps - sometimes more than once. 

This key lecture will discuss how the parts of the process work together.  Not to be missed! 

1 - Asking the right question Preview 02:48

What question are you asking?  What are your goals?
What does done look like? How good must our prediction be?

All these things are key parts of 1 - Asking the right question in the first place....

2 - Identifying, obtaining, and preparing the right data Preview 11:22

In this tell all lecture:

• Domain expertise reigns supreme! 
• Where will you get your data from?  Surprising secret sources of data you might not have considered.
• Dirty data....  dirty,dirty data!  Anticipating the largest effort in any Machine Learning  project realistically.. as well as discussing tidy data.

What are waiting for!  Go to your the lecture (room) and clean that (data) up!  All messes are not created equal. 

3 - Identifying and applying a ML Algorithm Preview 11:09

It's science and it's art.  In this lecture we'll discuss how Machine Learning algorithms interact with data to model answers to your problems.  We'll discuss and illustrate four common Machine Learning algorithms.  For each, we'll cover off how they work, and what workloads work best for them.  You'll become a master of the digitally arcane, with powers over:

• Decision Trees
• Naïve Bayesian Classifiers
• Neural Networks
• kNN - K's Nearest Neighbours.

4 - Evaluating the performance of the model and adjusting Preview 04:05

How do you evaluate the performance of your Machine Learning algorithm anyway?  And if it's not working they way you expected - how do you fix it?  In this tell all lecture, we'll discuss common problems of Machine Learning - and how address them.

5 - Using and presenting the model Preview 01:58

Finally!  We've reached the end goal!  Or have we?

In this brief lecture, we'll cover off four important things to keep in mind to use your Machine Learning Model. 

Machine Learning - Process

A quiz on the process of Machine Learning

How to apply Machine Learning for Data Science - Overview Preview 00:31

How do you get started in your journey to applying Machine Learning for Data Science?  In this brief overview, we'll describe the tell-all lectures, that will give you a place to start to apply Machine Learning and Data Science. 

Where to begin your journey Preview 00:58

HOW NOT TO LOOSE YOUR MIND.

Really.  This lecture is a important one, because it will give you guidance on how to get started in your journey without loosing your mind along the way. 

Common platforms and tools for Data Science Preview 01:57

What do you need to do Machine Learning?  Is it expensive?  Out of reach?

In this surprising lecture, we'll pull back the curtain on what Data Scientists are actually using.  We'll also list the top five tools for Data Science, that we will deep dive into, in the following lectures.

Data Science using - R Preview 01:57

The number one tool for Data Science, is "R" and is a power house for Machine Learning applications.  We'll describe the tool, as well as provide links an important tips on using it.

Data Science using - Python Preview 02:26

The second most popular tool for Data Science, is "Python".  Python is a general programming language with incredible power, versatility and flexibility.  It's gaining on R year by year, and has powerful Data Science and Machine Learning Capabilities. 

We'll describe the python, as well as provide links an important tips on using it.

Data Science using SQL Preview 01:50

The third most common tool for Data Science is SQL.  Pronounced SEA-QUEL, this is a Database language.  In this lecture we'll describe what SQL is, and why it has shown up in the third place for data science tools. 

Data Science using Excel Preview 01:44

The fourth most common tool for Data Science is Microsoft Excel?  Yes - really!  In this lecture we'll describe Microsoft Excel and it's value as a Data Science tool. 

Finally, we'll give you the "real deal", when it comes to doing Machine Learning in excel.  The answer, will surprise you!

Data Science using RapidMiner Preview 01:11

The final top five tool for Data Science is rapid miner.  In this lecture we'll discuss using Software as a Service, and some things to think about when using Rapid Miner. 

Cautionary Tales Preview 01:46

You made it!  In this final lecture of section 5, we'll talk about things to watch out for when doing Machine Learning.  This lecture will give you key information on how to avoid obstacles on your way to success!

All done! What's next? Preview 00:40

Congratulations on your journey into Machine Learning and Data Science.  We sincerely hope you enjoyed it - and we hope to see you again... in our next course! 

NOTE: November 2018 - The next course is *IN THIS COURSE*!  That's right - check out the next lecture for our included bonus course "Machine Learning in Python and Jupyter for Beginners"! 

Introduction and Anaconda Installation Preview 04:51

• Introductions!  Who am I?

• Who are you?

• Starting the Anaconda download.

What will we cover! Preview 06:12

• Prerequisite knowledge

• Topics for the course

• What won't we cover today?

• How the course will be delivered.

• Titanic survivability project - what we'll be building.

Introduction and Setup Preview 08:18

• Introducing Kaggle

• Where it the titanic example?

• Starting Anaconda Installation.

• Platform Selections -Why python?

• Platform Selections -Why python 3.x?

• Platform Selections -Why Anaconda?

Crash course in Python - Beginning concepts Preview 06:06

• Comments.

• Basic Variable and Assignments.

• Notes about Data Types.

• Data type Summary.

• Basic Type Casting.

• Advanced Assignments.

• Advanced Assignments - Error situations.

Crash course in Python - Strings, Slices and Lists! Preview 07:04

• Strings - Basic String Assignment.

• Strings - Unusual String Assignment.

• Strings - Basic String Operations.

• Strings - Core Concept - Immutability.

• Slices.

• Lists.

• Lists - Basic List Operations

• Lists - Additional Operations.

• Lists - Advanced Topics.

Crash course in Python - Expressions, Operators, Conditions and Loops Preview 05:41

• Notes about Expressions.

• Arithmetic and Bitwise Operators.

• Relational, Logical, and Identity Operators.

• Identity Operators.

• Assignment Operators and Membership Operators

• Conditional Logic and "if" statements.

• Iterations and Loops - Simple while loop.

• Iterations and Loops - Advanced loops and for loops.

Crash course in Python - Functions, Scope, Dictionaries and more! Preview 05:26

• Functions and variable Scope.

• Dictionaries.

• Dictionaries - Errors Situations.

• Dictionaries - Further Example.

• Getting Help!

• Further magic to explore - Where to go from here (to continue your learning on Python)

• Completing the Anaconda Installation

Hands on Running Python Preview 10:48

• Running Python Interactively.

• Running Python stand alone scripts.

• Running Python in Jupyter notebooks.

• How to use Jupyter:

  • Creating notebooks

  • Using notebooks

  • Saving notebooks

  • Types of cells

  • How the Kernel works (and how to manage it)

  • Getting help

• Help with Jupyter Markdown language.

  
  

Foundations of Machine Learning and Data Science - Definitions and concepts. Preview 05:38

• What is Data Science?

• What are Data Scientists?

• Data Science areas.

• What kinds of problems does Machine Learning Solve?

  • Classification

  • Regression

  • Clustering

• Can a Machine Learn?

Foundations of Machine Learning and Data Science - Machine Learning Workflow Preview 05:09

• What is Machine Learning? - Simplified Overview

• What does data look like?

• What does the data in our Titanic example look like?

• Types of Machine Learning:

  • Supervised

  • Unsupervised

  • Reinforcement Learning

• The 5 steps of a Machine Learning Workflow

Foundations of Machine Learning and Data Science - Algorithms, concepts and more Preview 07:09

• Example algorithms

  • Overview - Decision Trees

  • Overview - Naive Bayesian Classifiers

  • Overview - Neural Networks

  • Overview - kNN

• Evaluating the performance of a model and adjusting

  • Overfitting

  • Underfitting

• Further magic to explore - Where to go from where (to continue your learning).

Introducing the essential modules for Machine Learning, and NumPy Basics Preview 07:17

• Overview - Highest to lowest level.

• Overview:

  • SciKit-learn

  • SciPy

  • Matplotlib

  • Pandas

  • NumPy

• Basics of NumPy

  • Basic Creation and Assignments

  • Updating Values

  • Array Builders - Ones

  • Array Builders - Zeros

  • Array Builders - Choose your own

  • Matrices

• NumPy: Further Magic to explore - Where to go from here (to continue your learning)

Pandas and Matplotlib Preview 09:43

• Introducing Pandas - the Python Data Analysis Library

• Introducing Matplotlib - plotting library which produces publication quality figures in a variety of formats.

• Pandas:

  • Basic Series Creation and Assignments.

  • Basic Data Frame Creation and Assignments.

  • Creating a Data frame from CSV and reviewing it.

  • Exploring the Data - Data Shapes and Types.

  • Accessing and Changing the Data - Rows (cases) and Columns (features)

  • Removing Data

  • Filtering Data

  • Determining Unique Values

  • Simple Analysis

Analysis using Pandas, plotting in Matplotlib, intro to SciPy and Scikit-learn Preview 05:25

• Matplotlib:

  • Simple analysis and plotting

• Pandas:

  • Simple analysis and plotting

• Matplotlib - Further magic to explore - Where to go from where (to continue your learning).

• Pandas: - Further magic to explore - Where to go from where (to continue your learning).

• SciPy

  • The fundamental package for scientific computing with Python

  • Sparse matrix (example)

• SciPy: - Further magic to explore - Where to go from where (to continue your learning).

• Scikit-Learn:

  • Simple and efficient tools for data mining, data analysis, and Machine Learning

A Titanic Example - Getting our start. Preview 09:33

• Let's get our start by applying the 5 steps of Machine Learning Workflow to the titanic.

  1. Asking the right question.

  2. Identifying, obtaining, and preparing the right data.

  3. Identifying and applying a Machine Learning algorithm.

  4. Evaluating the performance of the model and adjusting

  5. Using and presenting the model.

• Step #1 - Asking the right question

  • Creating our Titanic Example file

  • Reviewing the data, and data dictionary

  • Importing out modules - Pandas, Numpy, Matplotlib, and Scikit-learn

  • Loading the dataframe

• Step #2 - Identifying, obtaining, and preparing the right data.

  • Reviewing the data, identifying gaps and problems with the data set.

A Titanic Example - Understanding the data set. Preview 11:25

• Step #2 - Identifying, obtaining, and preparing the right data.

  • Exploring the data with Pandas and Matplotlib - understanding people in the data set in terms of:

    • Survival of the disaster

    • Gender of people onboard

    • Age of passengers (histogram)

    • Classes of passengers

    • Age distribution in the Classes of passengers

    • Embarkation location

• Note: The goal of this lecture (and the next lecture), is to identify the right data and features to use in the Machine Learning algorithm.

A Titanic Example - Understanding the data set in regards to survival Preview 08:55

• Step #2 - Identifying, obtaining, and preparing the right data.

  • Exploring the data with Pandas and Matplotlib - understanding people in the data set in terms of:

    • Survival in relation to age (Scatter plot)

    • Survival in relation to gender

    • Survival in relation to passenger class

    • Survival in relation to passenger class and gender.

• Note: The goal of this lecture (and the previous lecture), is to identify the right data and features to use in the Machine Learning algorithm.

A Titanic Example - Preparing the right data and applying a basic algorithm Preview 12:57

• Step #2 - Identifying, obtaining, and preparing the right data.

  • Preparing the right data - adjusting gender.

  • Preparing the right data - filling in missing ages. 

• Applying a basic hypothesis:

  • Step #3 - Applying a algorithm (a basic one).

  • Step #4 - Evaluating the performance of the hypothesis and adjusting. 

• Applying Linear Regression:

  • Step #2 - Preparing the data (building the training features, and training target)

A Titanic Example - Applying regression algorithms. Preview 06:59

• Applying Linear Regression (continued)

  • Step #3 - Applying the algorithm (running fit)

  • Step #4 - Evaluating the performance of Linear Regression (Cross validation)

• Applying a polynomial regression

  • Step #3 - Applying the algorithm (running fit)

  • Step #4 - Evaluating the performance of Polynomial Regression (Cross validation)

A Titanic Example - Applying Decision Trees (example of overfit and underfit) Preview 06:00

• Applying Decision Trees:

  • Step #3 - Applying the algorithm (running fit)

  • Step #4 - Evaluating the performance of Decision tree (Cross validation)

  • What happened???  - Overfit!! Note: See resources in this lecture for the charts)

• Adjusting the algorithm

  • Step #3 - Applying the algorithm (running fit)

  • Step #4 - Evaluating the performance of Decision tree (Cross validation)

• Step #5 - Using and presenting the model.

• Conclusion of the Decision tree model.  What features did it decide are most important?

Conclusions - for our Titanic Example, important concepts and where to go next! Preview 06:06

In conclusion:

• Concept: "The algorithm with the most data selection wins!"

• Thoughts on:

  • Feature engineering

  • Data selection

  • Algorithm selection

• Further magic to explore - Where to go from here (to continue your learning)

  • Kaggle

  • Link to an amazing blog post

  • Links to several amazing Jupyter notebooks

• How to contact me!

• Thank you!

Bonus Article - The startling breakthrough in Machine Learning from 2016. Preview 03:09

Attached is an article I wrote, in early 2017 of one of the most important developments of 2016.  I think it's as relevant today as it was back then. 

I hope you enjoy it! It's included in HTML format, as well as attached in PDF.