Machine Learning

 Introduction

  • Artificial Intelligence – Beyond Arnold’s Terminators and Rajini ‘s Robots
  • Machine (is) Learning  – Knowledge is power
  • AI VS Data science VS ML  – What is what?

 Building blocks of ML

  • Data driven decision making  – Information is wealth
  • Understanding data: What is it telling us?
  • Prediction: What will happen?
  • Decision making: What to do?
  • Causal inference: Did it work?

 Understanding of data

Learn the basic characteristics of data sets and identify effective statistical tools and visualizations to glean insights from your data.
  • Ask the right questions about the data
  • Know which tools to use to unlock insights
  • How data visualization clarifies data

 

Concepts Covered

  • Visualizing One-dimensional & Multi-Dimensional Data
  • Using PCA, Clustering, K-Means, and Topic Models
  • Structured vs. Unstructured Data

 Making predictions

Understand the basic concept of linear regression and how it can be used with historical data to build models that can predict future outcomes.
  • How to build a model that fits best with your data
  • How to quantify the degree of your uncertainty
  • What to do when you don’t have enough data
  • What lies beyond linear regression
Concepts Covered
  • Supervised Learning
  • Regression
  • Classification
  • Neural Networks
Application: Using a provided data set, you’ll work through a scenario and answer application questions.

Making Predictions – Part 2

Classification is used to predict outcomes that fall into two or more categories, such as: male/ female, yes/no, or red/blue/green.
  • Compare the ability of different methods to minimize prediction errors
  • Make better predictions, based on your data and desired outcome
  • Use the right approaches to deal with data complexity
Application: Using the data set provided, predict the default risk of a pool of applicants in the credit card industry.
ai

Making Predictions – Part 3

Neural networks are much like the networks in the human brain. They are used in machine learning to model complex relationships between inputs and outputs and to find patterns in data.
  • What are neural networks and how do they work?
  • Explore the history and examples of simple and complex neural networks
  • How neural networks minimize errors, regardless of the size of your data set
Application: Working in the Tensorflow sandbox, you’ll complete an exercise on neural network architecture.

Decision Making

Decision making is about selecting the “optimal” decision or action in the presence of uncertainty, which all professionals face regularly.
  • Learn how the decisions you make impact the immediate future and beyond
  • Choose the right approach based on the environment, rate of information flow, and your goal
  • Strike the balance between exploration (identifying what we don’t yet know) with exploitation (using what we already know)

Concepts Covered

    • Optimal Decision in Presence of Uncertainty
    • Dynamic vs. Static Environment
    • High to Low Information Rate
    • Using Model Predictive Control, Markov
    • Decision Process, Multi-Armed Bandit, and Reinforcement Learning
Application: Retail planning, website design, and creating a perfect chess player.

Causal Inference

  • Cause and Effect Relationship
  • Thoughtful Experience Design
  • Randomized Control
  • Hypothesis Testing
  • Synthetic Control
  • Time-Series Forecasting

Final Hands-on Exercise

  • An implementation that covers all of what was discussed in this track