Introduction to Workflow for Data Science

Introduction, examples, use cases and tools used in data science projects.

Transcript

1. Introduction to Workflow for Data Science Santiago Lizardo 2019/12/11
2. • Sr. Engineering Manager • Hobbyist Data Scientist About the speaker 2
3. About the talk 3 • High level overview of the data science workflow • Use case in bug triage • Tooling for machine learning • Small demo
4. Data science workflow 4
5. Workflow definition 6
6. Data science workflow: 10,000 feet view Data Insights Actions 7
7. Data science workflow: 10,000 feet view Business Question Data Insights Answer Actions 8
8. Business question examples 9 • Predictions • Are we going to grow in more than X% next quarter? • Is this customer going to leave in the next 5 months? • Classification • Is this a server or a workstation? • What type of server this is? (web server, email, file, proxy, …) • Anomaly detection • Is this spike in uploads normal?
9. Workflow example 10
10. Workflow example 11
11. Workflow example 12
12. Workflow example 13
13. Workflow example 14
14. Workflow example 15
15. Workflow example 16
16. Workflow example 17
17. Workflow example 18
18. Workflow example 19
19. Industry standards • Knowledge discovery in databases (KDD) • Analytics Solutions Unified Method – DM (ASUM-DM) • Cross-Industry Standard Process – DM (CRISP-DM) • Team Data Science Process (TDSP) 20
20. KDD process 21
21. CRISP-DM workflow 22
22. ASUM-DM workflow 23
23. TDSP workflow 24
24. DS workflow v. simplified 25
25. Time spent on each step of the process 26
26. Time spent on each step of the process 27
27. Data wrangling • Data… • …sourcing • …exploration (EDA) • …cleaning • …imputation • …enrichment (Feature engineering) 28
28. Data wrangling | Data ingestion • Identify the data source(s) • Static dumps (CSV, TSV, Parquet, …) • Live databases (SQL, NoSQL, …) • Data scrapping (Web) • Database (SQL) • Exposing it to the data science project 29
29. Data and variable type 30
30. Model selection Supervised Model Unsupervised 31
31. Model selection Model Classification Supervised Regression Unsupervised 32
32. Model selection Supervised Model Unsupervised Clustering 33
33. Model selection 34
34. Model validation 35
35. Dissemination • Understandable by stakeholders • Digestible • Concise • Visual (charts, plots) • Accurate 42
36. Chart tools • Python • Matplotlib • Seaborn • Plotly (interactive) 43
37. Charts 44
38. USE CASE 45
39. Use case: Triage • Every morning new bugs are reviewed • Engineering team is identified and assigned • Done based on description among other attributes 46
40. Use case: Triage • Can it be programmed? • Too many rules • Use information about previous 30K tickets to classify new ones • Error margin is acceptable 47
41. Use case: Triage 48 • Jira ticket attributes to read from: • Summary • Components • Labels • Area • Repository • Jira ticket attribute to update: • Team
42. Use case: Triage 49 • High level plan • Extract data from Jira à CSV • Use Python and Pandas to read CSV and clean it up • Use NLP to extract the main info from text • Remove stopwords, punctuaction, etc • Use stemming to get to the root of words • Use XGBclassifier from scikit-learn • Hook Python script to ML algorithm
43. TOOLING 50
44. Tooling: Languages 51
45. Tooling • Local notebooks • Jupyter notebooks (.ipynb) • Cloud notebooks • Google Collaboratory • Amazon Sagemaker 52
46. Tooling: Deployment 53
47. DEMO 54
48. Closing thoughts • Data science is for the many not the few • Save data now for what you might want to ask in the future • Workflows are useful tools in projects • Data science can be hard, so can be some business requirements 55
49. Q&A 57