
Machine Learning in Python: 43 Algorithms Commonly Used
Udemy · English
About this course
- Language
- English
- Category
- Provider
- Udemy
- Data updated
You can already fit a model. What you cannot always do is say why that model and not one of the others — and that is the question you get asked. This course covers the 43 algorithms most commonly used in practice, one at a time, across 19 sections and about 19 hours. Each gets a full concept lecture and a separate hands-on demo with a seeded notebook you run yourself. Every number quoted in every lecture was produced by code that ships with the course. Where a result was surprising or unflattering, it is reported as measured rather than as expected. Every algorithm is introduced through the same five questions: what question does it answer, what does it need from you, what does it hand back, what does it cost to train and to score, and can you justify a single decision to somebody entitled to overrule you. By the end those cards stack into a comparison you can actually choose from. The 19 sections are organised by the question being answered rather than by mathematical family, so the table of contents is itself the first filter you apply. The nineteen sectionsOrientation: how to use this course · The shared vocabulary: what every algorithm assumes · How models learn: optimisation from the ground up · Predicting a number: fit a form, or copy your neighbours · Predicting a category: the four classical classifiers · Making it stronger: ensembles and gradient boosting · Finding groups: six ways to cluster, and how they disagree · Finding the unusual: anomaly detection without labels · Fewer columns, same signal: projection and embedding · What comes next: forecasting a time series · Seeing: convolutional networks for vision · Reading and writing language: recurrence, attention and transformers · Relationships and networks: learning on graphs · Creating data: adversarial and diffusion generation · Deciding under uncertainty: reinforcement learning · Knowing what you don't know: uncertainty and c…











