Introduction

Note: This post is for beginners who want to learn Machine Learning through python coding and for those who want a structured video lectures of the book, Hands-On Machine Learning with Scikit-learn, Keras and TensorFlow. For advanced users, I would suggest going through the curriculum in Notion for your own use case.

If you want a visual understanding then I would suggest following the short video on my youtube channel giving an overview of the series:

Hello World! I am Aditya Ramesh Ganti, a nerd-y (probably the word dosen’t exist, but nevermind!) student pursuing my B.Tech degree in Smart Manufacturing at PDPM Indian Institute of Information Technology, Design and Manufacturing, Jabalpur (wow, that’s a mouthful!😅 You can call it PDPM IIITDMJ). Jabalpur is a city in Madhya Pradesh, India.

I am a Machine Learning and Robotics Entusiast aiming to be one of the best in the field. I love doing research irrespective of the subject and gain knowledge, whether I remember it or not is secondary, though 😅. I have been learning Machine Learning and a little bit of Robotics for the past year and a half. And I thought I’d share my learning through the series of videos that I will be uploading weekly every Sunday starting today on my YouTube Channel. The series provides explanation of the concepts and do hands-on code provided in the book through Jupyter Notebook. All the code will be available in my github repo to get along with the videos. Though, the repo is still under development, all the code that were recorded will be provided before-hand.

Prerequisites

A pinch of Python and a touch of some familiar Python’s scientific libraries - Numpy, Pandas, and Matplotlib.

Also if you care about what is happening behind the scenes, you should have a reasonable understanding of college-level math as well. Khan Academy is a good place to brush up your math skills.

If you want a whole roadmap of what to learn to become a Machine Learning Engineer, I’d recommend the roadmap made by Mr. Daniel Bourke who is a self-learned Machine Learning engineer and is also one of my inspirations for creating content regarding the same. Thank you for all the hard work Mr. Bourke, looking forward to more videos of yours.

Why this course?

Actually I posted a discussion on his GitHub repo for the TensorFlow Deep Learning Course, regarding this series of videos…

image

And that’s how it went. And I plan on following what he suggested. I am going to inspire you to do research on your own. There is Google to answer our prayers!

JustGoogleItBroAllGasNoBrakesGIF

So this course is going to cover as much as it can but also give you extra-curriculum activities which you need to google. Of course, it is not compulsory but it is always beneficial to be out of the comfort zone and take the initiative. 😎

Hence, this course is going to be beneficial in not only gaining knowledge of Machine Learning, but also the methods of self-learning.

O’Reilly Hands-On ML Book Curriculum that we’ll be covering

Status

  • Finished chapter 1 June 3, 2022
  • Starting Chapter 2 June 3, 2022 1:04 PM

Part 1:

  • The Machine-Learning Landscape
  • End-to-End Machine Learning Project
  • Classification
  • Training Models
  • Support Vector Machines
  • Decision Trees
  • Ensemble Learning and Random Forests
  • Dimensionality Reduction
  • Unsupervised Learning Techniques

Part 2:

  • Introduction to Artificial Neural Networks with Keras
  • Training Deep Neural Networks
  • Custom Models and Training with TensorFlow
  • Loading and Pre-processing Data with TensorFlow
  • Deep Computer Vision Using Convolutional Neural Networks
  • Processing Sequences using RNNs and CNNs
  • Natural Language Processing with RNNs and Attention
  • Representation Learning and Generative Learning using Autoencoders and GANs
  • Reinforcement Learning
  • Training and Deploying TensorFlow Models at Scale

Part 3:

  • Machine Learning Project Checklist
  • SVM Dual Problem
  • Autodiff
  • Other Popular ANN Architectures
  • Special Data Structures
  • TensorFlow Graphs

The Machine Learning Landscape

Assign: Aditya Ramesh Ganti Date: June 2, 2022 13:00-15:00 Status: Completed - June 2022

1. What is Machine Learning?

  • Machine Learning is the science and art of programming a computers so that they can learn from data.

2. Why use Machine Learning?

  • It is difficult to write and maintain a traditional program for the problem of spam filter as the input varies every time.
  • Instead predicting through some given key words by a Machine Learning algorithm is much easier. Hence comes the use of machine learning.
  • Also, machine learning can help humans learn about the various possibilities that can occur in an event given the few dependent features.

To summarize, Machine Learning is great for:

  • Problems for which existing solutions require a lot of fine-tuning or long lists of rules: one Machine Learning algorithm can often simplify code and perform better than the traditional approach.
  • Complex Problem for which using a traditional approach yields no good solution: the Best Machine Learning techniques can perform better than the traditional approach.
  • Fluctuating environments: a Machine Learning system can adapt to new data.
  • Getting insights about complex problems and large amounts of data.

3. Examples of Applications

  1. Analyzing images of products on a production line to automatically classify them
  2. Detecting tumors in brain scans
  3. Automatically classifying news articles
  4. Detecting credit card fraud
  5. Building an intelligent bot for a game

and much more…

4. Types of Machine Learning Systems

There are so many different types of Machine Learning Systems that it is useful to classify them in broad categories, based on the following criteria:

  • Whether or not they are trained with human supervision
  • Whether or not they can learn incrementally on the fly.
  • Whether they work simply by comparing new data points, or instead by detecting patterns in the training data and building a predictive model, much like scientists do.

4.1 Supervised/Unsupervised Learning

Machine Learning systems can be classified according to the amount and type of supervision they get during training. There are four major categories: supervised learning, unsupervised learning, semi-supervised learning, and Reinforcement learning.

4.1.1 Supervised Learning

In supervised learning, the training set you feed to the algorithm includes the desired solutions, called labels.

Here are some of the most important supervised learning algorithms:

  • k-Nearest Neighbors
  • Linear Regression
  • Logistic Regression
  • Support Vector Machines (SVMs)
  • Decision Trees and Random Forests
  • Neural Networks

4.1.2 Unsupervised Learning

In unsupervised learning as you might have guessed, the training data is unlabeled. In simple words… The system tries to learn without a teacher.

Here are some of the most important unsupervised learning algorithms:

  • Clustering
    • K-Means
    • DBSCAN
    • Hierarchical Cluster Analysis (HCA)
  • Anomaly detection and novelty detection
    • One-class SVM
    • Isolation Forest
  • Association rule learning
    • Apriori
    • Eclat