Related Posts
XLM to 1.00. Pump it up!
Pro
Penn or draftkings?
Bowl Leader
Additional Posts in Data & Analytics Consultants
Got messaged by a C3 . ai recruiter. Read that wlb is bad and that the interview process is absurdly long, but the Glassdoor reviews are 4.2 and can't find actual hours worked posted by anyone. How's the culture really? I'd be aiming for DS consulting, something more functional but with DS/ML concepts as my differentiator.
C3.ai, Inc.
Mentor
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
Mentor
School: Ring Theory, Number Theory, Markov Chains, The NP complete assignment models from my PhD dissertation.
Work: Some if the time series regression models I’ve seen. Bayesian marketing models. [lightweight compared to above]
I’m lucky if I build a scatter chart in excel these days.
In school: Extractors and expanders in pseudorandomness. Still the hardest concept to understand I’ve ever come across.
On the job: the delta method to calculate a confidence interval for the odds ratio of an A/B test. Kind of hard to understand the math, but the calculations are easy.
Average day: I can calculate averages in Excel like a fiend.
School: Measure theory, time series proofs, explaining Bayesian multivariate time series procedures to an industry panel (never again).
Work: RL model convergence (unsuccessful ❤️)
Avg day: for a bit in there, the above RL stuff; how I learned real research may not be for me. Quaternion rotation, if only because multiple complex dimensions make people whoa. But mostly making fancier heuristic metrics than existing ones used by the business.
Lots of junk you can't take at face value gets published, sometimes you need to confirm yourself by at least retracing the math. Catching inconsistencies early may ward off running an expensive failed experiment or provide theoretical guarantees needed in cases when experimenting is prohibitive
School: Partial differential equations, singular value decomposition. Fourier transforms, Taylor series
Work: state space repenetation of dynamic systems, maximum likelihood estimation, Taylor series, Weibull analysis, general solutions to 4th order polynomials (which I solved numerically cause fuck that noise)
I really hate statistics though. I’d much rather stay in my deterministic bubble, the world just makes more sense here.