Prasanna Bhogale 12-12-2018
- Culture and organization : links with software engineering (version control, agile, high quality code)
- Tool mediated advances : algorithms closely linked to developments in hardware (GPUs) make whole new classes of problems tractable
- Data science in the wild : the cloud, the docker revolution and devops (you build it, you run it)
- Visualization : Javascript libraries and ever increasing expectations about quality and interactivity of visualizations
- Powerful conceptual frameworks : Auto diff, gradient descent (TF, PyTorch), causal graphs (Judea Pearl), Bayesian everything
- Exploring modern data science from the R ecosystem (and explore surroundings.. JS ? Stan ? Keras ? API building ?)
- Learning new concepts and skills that unlock important, interesting or beautiful aspects of data science in R
- Covering full stack data science in R - from statistics to deep learning in production and everything in between
- Exploring interesting use cases and domain specific challenges
- Welcoming : there are no stupid questions, there is always help available
- Uncool : make hard things easy, don't create inaccessible cliques around certain skillsets
- Curious : learn everything, become better at everything, explore application domains
- Useful : every talk should have
- clear conceptual takeaways
- reusable code
- Collaborative : Pitch in to give talks, do side projects, contribute to the R community
Prasanna Bhogale [email protected]