# Lecture 4 ## Topic In this lecture we look at the basics of machine learning and work with sklearn for the firs time. ## Lecture Slides ```{raw} html ``` ```{eval-rst} :download:`Download the slides <_static/bld/pdfs/lecture_4.pdf>` ``` ## Exercises ```{toctree} --- maxdepth: 1 --- bld/notebooks/exercises/exercise_4.ipynb bld/notebooks/solutions/exercise_4.ipynb ``` ## Suggested Homework - Re-do or finish the exercises from the class - Do hyperparameter-tuning for logistic regression, random forrests and boosting - Experiment with other datasets ## Additional materials ### Introduction to Machine Learning with Python A slightly outdated but very accessible introduction to machine learning is the book [Introduction to Machine Learning with Python](https://www.oreilly.com/library/view/introduction-to-machine/9781449369880/) by Andreas Müller and Sarah Guido. In the lecture we had only time to cover supervised learning (and there only classification). The book would be a good source for anyone who wants to read up on unsupervised learning methods (e.g. clustering and dimensionality reduction). It also contains a lot more background on the things we covered here and ### Chapter 5 of the Python Data Science Handbook We have used the [book](https://jakevdp.github.io/PythonDataScienceHandbook/05.00-machine-learning.html) before. It starts by introducing some fundamental concepts and continues with in-depth chapters for different methods. Definitely have a look if you want to use a method we did not cover in enough detail here. ### JEP Paper This [paper](https://www.aeaweb.org/articles?id=10.1257/jep.31.2.87) by Mullainathan and Spiess (Journal of Economic Perspectives) is a great summary of the similarities and differences between econometrics and Machine Learning. ### Online-courses If you want to get much deeper into machine learning, you can have a look at Sebastian Raschka's [online course](https://sebastianraschka.com/blog/2021/ml-course.html). He also has a very recent course on deep learning. ### Summary paper about scores for multiclass classification Comprehensive paper [paper](https://www.arxiv-vanity.com/papers/2008.05756/#:~:text=Accuracy%20is%20one%20of%20the,computed%20from%20the%20confusion%20matrix.&text=The%20formula%20of%20the%20Accuracy,confusion%20matrix%20at%20the%20denominator.) on the different scores for multiclass classification.