Introduction to Machine Learning

Introduction

Some background
Looking 5-10 years ahead
What is Learning?
What is Intelligence?

Fundamental concepts

Applications
Tools and Libraries – Scikit Learn, matplotlib, Pandas, Numpy
Data Sources
Data Pipelines
Code along projects
Projects to work on
Primitives (if you want to learn from scratch))
What we won’t cover – Deep Learning and NNs

Sessions:

Day-1

1. Introduction and high level concepts (the big picture)
2. Code walk through of 2 applications
– a supervised learning app – predicting house prices
– an unsupervised learning app – clustering (u(using k-means)
3. Hands on coding session (for your projects))
4. Day-1: Wrapup – An interactive session on what you learned and what you would like to know.

Day-2

1. Applications of ML in the industry and how they use various ML techniques
2. An introduction to Text Analysis
3. A list of primitives you can code
4. Hands-on session (Customized)
5. Continue working on the projects
6. New topics to explore (on your own)
7. Resources for continuing self-learning
8. Wrap up

Fundamental Concepts

  • What is Machine Learning?
  • What is Data Science?
  • What is AI?
  • What is Data Engineering?

Machine Learning

  • Supervised Learning
    • – prediction, classification
  • Unsupervised Learning – discovering hidden structures
    – clustering
  • Reinforcement Learning – self learning robots

A road map to building ML apps

  • What do you do with ML?
  • Building intelligent applications (data to knowledge)
  • Prediction (based on past experience)
  • Regression
  • Classification

How to build an Supervised learning app?

1. What is the question
2. What will the answer look like
3. What is the data available
4. Cleaning and normalizing data
5. Identifying features (feature engineering)
6. Splitting data into training and testing
7. Picking an algorithm
8. Training the app – generating a model
9. Testing the app with test data
10. Hosting the app

Examples to work on (will be customized based on audience needs)

– Data sets
– Problems
– Approach