Data Science Methodologies: Making Business Sense
1h 51mIntermediate2024-12-12
Authors

Neelam Dwivedi
Assistant Teaching Professor at Heinz College
Course details
There is an increasing recognition that data science needs to go beyond small-scale experimentation to a large-scale implementation. In this course, Neelam Dwivedi brings software engineering and data mining methodologies to data scientists, then applies these ideas by taking a simple business need through an entire life cycle—hosting a model, consuming it in a web application, and setting up its CI/CD pipeline. Neelam begins by explaining the methodologies used in the course and how they are combined. She shows you where to begin in developing architecture and deploying a model, then explains how larger web applications may consume the model as a service. Neelam covers how to stage your model and the app, as well as how to plan ahead with an overall roadmap. She concludes with thoughts on how to further applications of data science methodologies.
Skills covered
Data ModelingData Science FoundationsData ScienceDeep Dive (X:Y)
Concepts
0. Introduction
- 01 - Models and the real world
- 02 - What you should know
1. Methodologies for Data Scientists
- 03 - Why methodologies
- 04 - Data science vs. software engineering
- 05 - Data mining methodologies
- 06 - Software engineering methodologies in data science
2. Deploy and Consume a Model
- 07 - Where to begin
- 08 - Development architecture
- 09 - Deploy the model
- 10 - Consume the model
- 11 - Challenge 1
- 12 - Solution 1
3. Take a Model from Dev to Staging
- 13 - Set up a CI CD pipeline
- 14 - Deploy the app
- 15 - The roadmap
- 16 - Challenge 2
- 17 - Solution 2
Conclusion
- 18 - Furthering applications of data science methodologies