We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Thomas Wiecki: The State of the Art for Probabilistic Programming | PyData Global 2022
Learn about probabilistic programming advancements with Thomas Wiecki at PyData Global 2022. Explore PyMC's modularity, tech mixing, community, and tools. Discover innovations like quantum computing backends.
- Thomas Wiecki discussed the state of probabilistic programming in his PyData Global 2022 talk.
- PyMC, a popular probabilistic programming library, is under active development and shows promise in its recent advancements.
- PyMC’s flexibility in incorporating structures like seasonality and crafting kernels is a strong point.
- PyMC’s use of different backends, such as JAX, allows for a modular, high-level model description while still offering lower-level optimizations.
- The unifying layer of SARA/PyTensor allows for mixing and matching different technologies, creating an ecosystem for innovative Bayesian modeling approaches.
- PyMC’s user-friendly interface and extensive documentation make it an excellent tool for teaching and learning probabilistic programming.
- Plotting capabilities and meetups, like the PyCausal conference, help foster a vibrant PyMC community.
- Innovations in samplers, such as BlackJAX, offer exciting possibilities for Bayesian inference.
- Quantum computing backends are an experimental but promising avenue for future development.
- Though GitHub stars don’t tell the whole story, the growing interest in PyMC suggests its rising popularity in probabilistic programming.
- PyMC’s capacity to handle complex, real-world problems while maintaining ease of use makes it a powerful tool for industries like marketing and finance.