Problem Identification and Solution Design for Data Scientists

Course details

Whether you’re working as a consultant or an employee, you need to be able to speak to nontechnical business leaders if you want to be successful as a data scientist. Becoming bilingual isn’t easy, though—so what’s the secret to mastering this coveted skill? In this course for aspiring data professionals, join instructor Keith McCormick as he outlines the fundamentals of problem identification and solution design for data scientists. Learn how to translate business needs into technical terms, follow the CRoss Industry Standard Process for Data Mining (CRISP-DM), conduct structured interviews with project leaders, and think about projects from a business leader’s point of view. Keith offers quick insights and easy-to-use tips for conveying technical concepts clearly with ease. Along the way, you’ll learn how to leverage business communication and documentation skills throughout every stage of a project.

Learning objectives
Identify the tasks associated with the business understanding phase of the CRoss Industry Standard Process for Data Mining (CRISP-DM).
Translate business goals into technical goals that can be achieved with data science techniques.
Identify project success metrics that align with a project sponsor’s business goals.
Diagnose possible causes of the challenges that a project seeks to overcome.
Explain options and technical considerations that a project sponsor might have to consider in business terms.

Concepts

0. Introduction

01 - Welcome
02 - Intended Audience

1. Meet the Project Customer

03 - Looking at the project from the sponsor's point of view
04 - Reading the project description critically
05 - Preparing for the first meeting as an internal resource
06 - Preparing for the first meeting as an external resource

2. What to Expect in the Initial Meeting

07 - What to expect in the initial meeting
08 - Inference vs. prediction
09 - Prediction vs. forecasting
10 - Avoiding confusion with other analytic project types
11 - Diving deeper into predictive analytics
12 - Identifying ROI
13 - Some questions that always apply

3. The Business Understanding Phase of CRISP-DM

14 - Business understanding phase overview
15 - The four tasks
16 - Advice on how much detail to share

4. Meeting with IT

17 - Pre-meeting requests and preparation
18 - Ensuring that you leave with the critical info
19 - The provenance of the data
20 - Leveraging IT experience and expertise

5. Meeting with SMEs and the Frontline Team

21 - Leave these interviews for last
22 - Assessing culture
23 - Focusing on inclusion, not exclusion

6. Avoid Delays

24 - Metadata vs. data
25 - Estimates of accuracy
26 - Beware too much prototyping

7. Writing Up Your Results

27 - Documentation advice
28 - A simple visuatlization for setting priorities
29 - Striking the optimal amount of detail

Conclusion

30 - What's next in problem identification and solution design for data scientists