Suggestions
Artur Suchwalko, PhD
Chief Data Science Officer & Co-Founder at QuantUp ✹ Unlocking possibilities: AI solutions for e-commerce, life sciences, fintech, retail, logistics, and telecoms
Artur Suchwalko, PhD, is a prominent figure in the field of data science and advanced analytics. He serves as the Chief Data Science Officer and Co-Founder at PrediqTank, where he utilizes his extensive expertise to lead various projects. In addition to his role at PrediqTank, he is also the CTO and Co-Founder of QuantUp, a company focused on leveraging advanced analytics and machine learning.
Dr. Suchwalko has over 20 years of professional experience and has successfully led more than 100 projects across different industries. He holds a PhD in mathematics, which underpins his analytical capabilities and contributions to the field.123
Highlights
We cannot work on AI solution now.
We will come back after the data warehouse is built.
Will you wait for the perfect data?
It will never happen!
My advice:
🔵 State the business problem precisely. 🔵 Take a dataset that you can have now or soon. 🔵 Check its accuracy, completeness, and analyze the processes that generated it. 🔵 Check what you can do with it. 🔵 Try to estimate the financial outcome. 🔵 If it's good enough, there will be an appetite for more.
#datascience #machinelearning #ai
Q: What is AUC?
A: I don’t know precisely. I cannot explain.
Q: Knowing that this is [...] how would you use it?
A: I don’t know exactly.
These are questions and answers from real job interviews with data scientists with some experience.
Would you hire someone who answers this way?
There are several levels of understanding of ROC and AUC:
🟠 Intuition behind these concepts 🟠 Applications of them 🟠 How to use it? 🟠 Why AUC is often chosen instead of the other metrics? 🟠 Properties of ROC and AUC, pros and cons of using hem 🟠 Precise definitions 🟠 Relationships of AUC to the other measures 🟠 How is AUC related to the optimization criteria used in ML algorithms?
I prefer to work with people who understand important data science concepts at all these levels. The whole QuantUp team consists of such people.
Is such a deep understanding necessary for successful data science projects?
Not always.
It is necessary if you’d like to build excellent solutions.
#datascience #machinelearning #ai
