Abstract: Data science improves effectiveness of many industries by looking at what has occurred in the past and using that analysis to help make decisions about the future. This talk shows practical techniques to solve questions about Agile processes and software development using even small amounts of historical data.
This session introduces the concepts behind data science and offers numerous easy wins with practical applications to any software development process. It will demonstrate how even a little data can be used to inform more likely future outcomes and how to get started immediately in your company.
Some of the practical techniques explained and demonstrated are –
- How to estimate the likely lead time for future items based on (similar) past items using completed item cycle-time data trends
- How to forecast how many items are likely to be delivered over some period of time using system throughput (just needing historical start and completion date data)
- How to estimate the likely number of remaining defects in a product using sampling and defect report and fix rate data
- How to look for clusters of similar impediments and failures in completed items using blocker clustering and frequency data
- How to identify and quantify declining predictability in a process earlier by observing changing process trends
By the end of this session you will know how and why simple techniques applied to historical data are reliable and outperform intuition alone, and have immediately actionable techniques that you will understand. Everything shown is easily implemented using post-it notes and spreadsheets (yours or mine, freely available on Google sheets or Excel).
Learning outcomes include -
- Learn what “data science” means and how it is used with simple examples
- Learn how to immediately start doing simple analysis of historical data
- Learn how to assess the expected reliability of analysis applied to historical data
- Learn five immediate ways to perform quality data science in an Agile context
Data science isn’t as complicated as it sounds with the core concepts easily understood in a few minutes. Even if you hated mathematics in school, this session will make you love again.
Learning Outcomes: - What “data science” means and how it is used with simple examples
- How to immediately start doing simple analysis of historical data
- How to assess the expected reliability of analysis applied to historical data
- Five immediate ways to perform quality data science in an Agile context
Attachments:
