You will get to know
✅ why visualisation is your friend and you will learn simple tips on how to do it and which packages are most commonly used (matplotlib, seaborn)
✅ why success metrics are a 'beacon' in machine learning
✅ which success metrics are most commonly used and for which types of models (classification and regression)
✅ why model validation is so important, you will learn about the types including cross validation
✅ what is underfitting vs. overfitting, and how to diagnose overfitting using a learning curve
✅ you will learn more ML models using the sklearn library, this time focusing on tree-based algorithms including bagging: (DecisionTreeClassifier, RandomForestClassifier, ExtraTreesClassifier)
✅ user-friendly dynamic way to change the content of graphs with ipywidgets, without having to create a large number of the same type of lines of code
🎁 bonus (review of one of the useful Python packages - plotly)
Tasks
1️⃣ binary classification: forecasting earnings (over 50,000 or less)
2️⃣ multi-label classification: forecasting land type
Result
After going through this module, you will learn how to correctly interpret model evaluation metrics in order to make the right decisions. And you can immediately apply your visualisation knowledge to any of your data, without having to buy or install other software.