About the talk
In today's world, machine learning has been used extensively in our daily life, from sending over a text to browsing on social media. To understand machine learning, Gradient Boosting algorithm would be a great starting point. In this talk, I will code and explain everything from scratch: from building of an individual Decision Tree to the ensembling logic. We will also compare the performance of our implementation to the industry standard library — sklearn. In a few lines of code viewers will see how Machine Learning works under the hood.
This talk will cover
- Numpy: a Python library for fast array manipulations which we will need to understand the code
- Machine Learning: brief introduction and testing a baseline solution on a toy dataset
- Decision Trees: how they learn from data and make predictions
- Gradient Boosting: the mechanism of ensembling individual Decision Trees to improve the model's performance
Programming & Development
About the speaker
Grigorii likes to extract value from big amounts of semi-structured data.