Sivan Aldor-Noiman, VP of Data Science at Wellio, presented “The Past/Present/Future + Myths of Data Science” at Domino. This blog post provides a few highlights from the interactive talk as well as the full video.
At Domino HQ, Sivan Aldor-Noiman presented insights and led an interactive dialogue about the past, present, and future of data science. Aldor-Noiman discussed core components of data science that include: products, people, process and platforms. A few key highlights from the talk, “The Past/Present/Future + Myths of Data Science”, include
- Products — The first industries to adopt Data Science were finance, marketing and search engines (e.g. Google). This was mostly driven by the vast data these industries were able to accumulate. Nowadays, due to the advancement in technology (e.g. IoT) data is becoming more available in new area such as agriculture, supply chain management and transportation. In the future Aldor-Noiman hopes that Data Science will be more integral in places such as public education where she believes Data Science will be able to help each child learn in a manner that will invigorate them and help them reach their real personalized potential.
- People — Moving away from distinct roles. Even five years ago, there was a big distinction between software engineers , that were writing production code, versus Data Scientists that were hired to create metrics and evaluate them. In the present day, there is more overlap between roles. Roles are also impacted by size of the organization. While some of the larger organizations can afford to have people specialize, other smaller organizations are looking for people that can support the CI/CD process, i.e. “who can take models, put them in production, evaluate themselves, evaluate the models, be more self-sufficient”. Academia is also starting to support a more holistic education to both software engineers and statisticians, allowing them to become a lot more valuable in industry. In the future, Aldor-Noiman believes we will see more people with coding skills and statistical critical thinking.
- Process — Data scientists are becoming more involved in work flows beyond data discovery and model development. Data Science is adopting the engineering organization CI/CD framework and as such Data Scientist are getting more involved throughout the process from product and design definitions all the way to monitoring model performance in production.
- Platform — In the past it was not uncommon for data scientists to question whether the model deployed into production was the model they developed. Some organization were forced to also develop their own private Data Science platform to improve Data Science workflows. Today, she believes that there are many good solutions for Data Science platforms that no longer requires organizations to build their own Data Science platform. Aldor-Noiman advocates for an ideal future state where a data scientist has additional visibility into model development, model deployment and model monitoring.
- Myths — Aldor-Noiman wraps up the talk by delving into data science myths including “More data, better models”, “Best model always wins”, “AI is not data science”, and more.
If interested in additional insights from this interactive talk, the full video is available for review.
New to Domino? Consider a Guided Tour.Start Your Free Domino Trial
Recent PostsTensorflow, PyTorch or Keras for Deep Learning Reinforcement Learning: The K-armed bandit problem KNN with Examples in Python Computer Vision in Deep Learning: An Introductory Guide Supervised vs. Unsupervised Learning: What’s the Difference? Powering Up Machine Learning with GPUs A Guide to Natural Language Processing for Text and Speech What Is Reinforcement Learning and How Is It Used? A Detailed Guide To Transfer Learning and How It Works Introduction to Deep Learning and Neural Networks
Other posts you might be interested in
Subscribe to the Data Science Blog
Receive data science tips and tutorials from leading Data Scientists right to your inbox.