I'm taking the year off from conference engagements due to the COVID19 pandemic. I look forward to speaking again in 2021 or 2022 when the global health situation has improved.
Introduces the concepts of outlier analysis and anomaly detection for an audience of data scientists or analysts, explaining how alerts may be used in operations and how to implement simple yet reliable methods for alerting with acceptable false positive rates.
play in browser
Gives an overview of life cycle stages for data science. Tell which prerequisite stages you have completed and know what to do next. Managers learn which steps can be contracted out and how to hire; job seekers learn how to tell if a prospective company is really ready to do what they say they're going to do, and how to learn and practice for the specific type of work you want to be doing. I offer a taxonomy of what specifically is accomplished in each stage as an alternative to ambiguous job titles.
play in browser
Describes a organized process for investigating systems which used to work and now don't, with particular applications to web technology stacks. Especially helpful for computer scientists who are not familiar with system engineering and understanding emergent behavior; good training for developers going into operational roles. Shows that some data you can collect will usefully partition your possible problem space, and other data is “associated” with the problem but not actually useful. Also includes guidance on how to teach this skill to the junior members of your team.
slides (sadly, I do not have audio or video for this one)
How does data science differ from software development? What are appropriate ways to run an agile data science team? Discusses prioritization of projects by estimating benefits assuming a successful technical completion, testing a minimal implementation as quickly as possible, and working with the business and engineering.
To watch and listen, open the audio first, then come back here and open the slides while the audio plays.
As an entrepreneur or executive at a company which has no existing data science function, how can you hire your first person and get a successful outcome? How to understand your needs and goals. Also goes into the use of a take-home exercise and presentation for hiring in more detail than my 2018 talk.
slides video slides
I'd be happy to discuss creating a talk on your preferred topic for your conference or gathering.
Contact Terran About Speaking Read More Articles
If you have somehow arrived here looking for my old work on US airspace congestion or running Hadoop clusters, let me know and I'll do my best to find it for you in the archives.