Benjamin Wee

Mostly Bayesian stats, #Rstats, #Python | Currently working on climate policy/modelling

[Benjamin Wee @BenwWee]

Super happy to post my Master's research on using Simulation Based Calibration (SBC) to compare MCMC algorithms in the context of stochastic volatility (SV) models https://t.co/DD72Da6JsR. I hope to write a blog post to summarise, but in the mean time I'll do my best here 🧵:

[Lauren Kennedy @jazzystats]

My master's student @BenwWee just posted his research to Arxiv https://t.co/iV6PeoUEIq!
In it he compares two methods to estimate stochastic volatility models.
I really loved thinking about challenges and ideas for fairly comparing samplers while working with him.

[Benjamin Wee @BenwWee]

@MattCowgill @economeager @scepticalranil 😮 I didn't know about this - I just subscribed!

[Benjamin Wee @BenwWee]

Hi @adamliaw, what brands of carbon steel pans/skillets do you use or recommend?

[Adam Liaw @adamliaw]

Before you ask, all my frying pans are carbon steel, except for one cast iron, and one non-stick I use for eggs and pancakes.

[Benjamin Wee @BenwWee]

@innesbigaran Morning coffee in a room with natural light, all devices away or in another room

[Benjamin Wee @BenwWee]

Referenced papers can be found here.

Talts et al (2020): https://t.co/tfqpILUNBR
Kim et al (1998): https://t.co/6Iwvc2MnS7

/end 🧵

[Benjamin Wee @BenwWee]

Super happy to post my Master's research on using Simulation Based Calibration (SBC) to compare MCMC algorithms in the context of stochastic volatility (SV) models https://t.co/DD72Da6JsR. I hope to write a blog post to summarise, but in the mean time I'll do my best here 🧵:

[Lauren Kennedy @jazzystats]

My master's student @BenwWee just posted his research to Arxiv https://t.co/iV6PeoUEIq!
In it he compares two methods to estimate stochastic volatility models.
I really loved thinking about challenges and ideas for fairly comparing samplers while working with him.

[Benjamin Wee @BenwWee]

I omitted many details in this thread, but this hopefully gives a good overview of my research. I had such a great time working with @jazzystats and Prof Catherine Forbes - I'll always be grateful for the time they invested in me and everything I've learned from their mentorship.

[Benjamin Wee @BenwWee]

Been thinking about how to structure data science code and how to reason about when to use OOP vs functional style coding for different parts of workflow. Thankfully, @ericmjl has also thought about this and written it all down!

https://t.co/tCuxyOuGtg

https://t.co/XQShW1tsTO

[Swen Kuh @swenkuh]

Festive season is over but this still feels like a gift, my first published paper on Statistics in Medicine (https://t.co/oEPEiaOZGx) ! In it we looked at what happens when you use leave-one-out (LOO) cross validation to validate MRP models for population inference.

[Benjamin Wee @BenwWee]

Chuffed to finally meet @hugobowne in person! I listened to Hugo's podcast 6 years ago when I started to get into data science and loved listening to him speak in person today. Thanks for the insightful chats, next beer on me! 🍻