Here we're trying to implement, check, or examine some theories, suggestions or software tools (libs). All by ourselves, this code is not from some courses teaching materials.
Well Ok, there's no 'we', all of those are my one's, but I will follow the scientific\educational works' approach when all the actions of implementer(s) are described in plural form.
| Program's name | Description | Used libs |
|---|---|---|
| "Standard_scaler" realization | Scaling is quite simple concept: you normalize your distribution into some scale (basic scaler you're easily think up to is dividing the whole set by abs max). It's needed in linear models and models that uses distances between vectors. Though do we understand its implementation? | pandas, numpy, seaborn |
| kNearestNeighbors realization | k-NearestNeighbors (KNN) is one of the easiest to understand ML-algorithms. But could we implement this thing based on theory from scratch? | pandas, numpy, matplotlib |
| ROC curve and ROC-AUC | We will calculate data for Receiver Operating Characteristics curve, then plot it and then we'll try to calculate area under that curve. | pandas, numpy, matplotlib, sklearn |
| Most frequent number (not algo) | We're going to explore several methods of finding the most frequent element in a set - quite a common task. The goal is to determine, which one actually works fastest. Don't be confused, this is not an algorithm task for finding most frequent object. | pandas, numpy |
just one of the pics I generated on March, 8th for datascience girls. nice, right?