PyData Meetup Recap
They say it takes a village to build and grow an engaged community.
We were privileged to host the monthly PyData meetup of PyDataSofia and Data Science Society groups and welcome the local data community last week. In this blog post, we’ll recap some highlights from the meetup and share some useful resources.
Campus X, where our office is also situated, were our host and the evening flew by with the buzz of the ecosystem networking, the energy of connections was forged and fuelled by delicious tidbits and drinks. In the spotlight was the "Bias-variance trade-off" topic from the highly skilled expert in the field, Georgi Nalbantov. Ph.D.
Georgi has 80+ publications in renowned journals in the fields of Computer Science, Healthcare, Finance, and Statistics. Co-owner of a patent in DNA analysis, his interests are in banking, predictive maintenance, ECG signal analysis, inventory optimization, and computer vision. In his talk, Georgi Nalbantov debunked one of the most important and tricky issues in machine learning, the concept of bias-variance trade-off.
The bias-variance trade-off is a design consideration when training the machine learning model. Certain algorithms inherently have a high bias and low variance and vice-versa.
So why learn more about it? Understanding the concept will help you make an informed decision when training your ML models.
As an experienced lecturer, he opened up with some data science jokes.
Below is the famous graph from the book of Hastie, Tibshirani, and Friedman - "Elements of statistical learning", which depicts the training and test errors. "As the model complexity goes up, the training errors go down, the test errors initially go down, then go up. We look for the complexity level at which the test error is minimal, which is called the best trade-off between bias and variance."
Georgi then provided various examples to illustrate the concept's application and keep the audience interested throughout the talk. Georgi recorded all of the examples he showed in a video on his YouTube channel. ▶️ You can watch it here ◀️.