Learn filter, wrapper, and embedded methods, recursive feature elimination, exhaustive search, feature shuffling & more.
InstructorSoledad Galli
The curriculum covers a wide range of selection categories including Filter, Wrapper, Embedded, and Hybrid methods.
Editorial course preview
These 4 complementary views highlight concrete, legible examples from the course presentation.
This educational slide defines feature selection as the process of choosing relevant variables for machine learning model building, illustrated by a diagram of features flowing into a central process.
This diagram illustrates the diverse sources used for feature selection, including Kaggle competitions, scientific articles, business practice, Scikit Learn, and the internet.
This slide outlines the course audience and difficulty levels, targeting data scientists, analysts, and software developers across beginner, intermediate, and advanced skill tiers.
This slide outlines key course benefits including over eight hours of video lectures, lifetime access to materials, ready-to-use code examples, and global insights.
Instruction includes specific statistical measures such as Mutual Information, Chi-square, and Anova for feature evaluation.
Learner signals indicate a desire for more dedicated code-along notebook structures rather than standalone lectures.
One signal suggests that some links to GitHub resources may be non-functional.
This course is best suited for intermediate practitioners who want to understand the mathematical and algorithmic logic behind variable selection to build more reliable models.
Score breakdown
The curriculum covers diverse categories including filter, wrapper, embedded, and hybrid methods with specific statistical tests.
While the course includes hands-on elements, signals suggest a need for more structured code-along notebook experiences.
Learners report clear and direct explanations, though some find the code complexity challenging without a programming background.
The course is updated regularly, though one signal noted issues with external GitHub resource links.
The curriculum aligns well with the technical requirements of data science, though it assumes a level of programming proficiency.