Machine learning is used in many different industries and fields. It can fundamentally improve the business if applied correctly. This chapter outlines machine learning use cases, job roles and how they fit in the data needs pyramid.

Machine learning and data pyramid

Terminology clarification

Order data pyramid needs

Match tasks in data pyramid

Machine learning principles

Modeling types

Find supervised and unsupervised cases

Job roles, tools and technologies

Job role responsibilities

Match data projects with job roles

Team structure types

Machine learning and data use cases

This chapter overviews different machine learning types. We will look into differences between causal and prediction models, explore supervised and unsupervised learning, and finally understand the sub-types of supervised learning: classification and regression. 

Prediction vs. inference dilemma

Inference and prediction differences

Identify inference vs. prediction use cases

Inference (causal) models

Experiments and causal models

Identify non actionable variables

Prediction models (supervised learning)

Supervised modeling principles

Identify classification and regression models

Prediction models (unsupervised learning)

Unsupervised modeling use cases

Classification, regression or unsupervised models

Machine learning types

This chapter reviews key steps in scoping out business requirements, identifying and sizing machine learning opportunities, assessing the model performance, and identifying any performance risks in the process.

Business requirements

Identify situation, opportunity and action

Identify successful experiments

Model training

Model training process

Training, validation and test

Model performance measurement

Poor performance examples

Identify performance metrics

Machine learning risks

Fixing non performing models

Non-actionable models

Identify actionable recommendations

Business requirements and model design

This chapter will look into the best and worst practices of managing machine learning projects. We will identify most common machine learning mistakes, learn how to manage communication between the business and ML teams and finally address the challenges when deploying machine learning models to production.

Machine learning mistakes

Identify machine learning mistakes

Data needs pyramid

Options

Not enough data

Late testing, no impact

Feature selection

Match ML mistakes by their types

Communication management

Business communication focus

ML market testing process

Market testing

Machine learning in production

Production systems

Machine learning use cases in production

Customer relationship management

Fraud prevention system

Autonomous car

Production systems ML use cases

Data Science Process

ML in production launch

Wrap-up

Managing machine learning projects

This course will introduce the key elements of machine learning to the business leaders. We will focus on the key insights and base practices how to structure business questions as modeling projects with the machine learning teams. You will understand the different types of models, what kind of business questions they help answer, or what kind of opportunities they can uncover, also learn to identify situations where machine learning should NOT be applied, which is equally important. You will understand the difference between inference and prediction, predicting probability and amounts, and how using unsupervised learning can help build meaningful customer segmentation strategy.

<h2> Learn the Basics of Machine Learning </h2>
<br>
This course will introduce the key elements of machine learning to the business leaders. We will focus on the key insights and base practices how to structure business questions as modeling projects with the machine learning teams. 
<br><br>
<h2> Dive into the Model Specifics </h2>
<br>
You will understand the different types of models, what kind of business questions they help answer, or what kind of opportunities they can uncover, also learn to identify situations where machine learning should NOT be applied, which is equally important. You will understand the difference between inference and prediction, predicting probability and amounts, and how using unsupervised learning can help build meaningful customer segmentation strategy.

Understand the fundamentals of Machine Learning and how it's applied in the business world.

Machine Learning for Business

Data Scientist in Python

Data Scientist in R

The team has reviewed the data sources: the systems are not capturing the data correctly and cannot be used in building the model.

The team has reviewed customers data: 3% of customers have data captured in a structured format, the rest is sitting in the back-end systems.

The team spent 2 months in building the prototype model. They need 8 more months to perfect it before running the A/B test in the market.

The model has been tested in the market and has not shown any results, the team says that they need another 6 months to perfect the model algorithm.

The team said they're going to use all data possible for the inferential model and that they don't need an input from the business team.

The team wants to get all the possible data before building a prototype. They need 8 months to aggregate data from all different sources.

Match ML mistakes by their types

Machine Learning for Business

Hands-on interactive exercise