Complete Bootcamp 2021 : Feature selection using Python

A Course by Kaggle grandmaster on Feature Selection : Machine Learning, Scikit Learn, Pandas, mlextend, clean your data

Feature selection is one of most important activity in machine learning/Artificial Intelligence pipeline. We select all relevant features for machine learning algorithm and discard less relevant or not relevant features. Feature selection is also known as  variable selection.This course will provide learner, detailed knowledge of feature selection. It is one of most detailed online course on feature selection.

What you’ll learn

  • Feature Selection using Python machine learning packages Pandas, scikit-learn(sklearn), mlxtend.
  • Learn the concept behind feature selection, detail discussion on feature selection method (filter, wrapper and embedded).
  • Filter methods selector like variance, F-Score, Mutual Information etc…
  • Wrapper Method : Exhaustive, Forward and Backward Selection.
  • Embedded Method : Lasso Decision Tree, Random Forest, ExtraTree etc.
  • Implemented with more than 15 Projects.
  • Ready to use code in machine learning projects.
  • Feature selection technique people used in Competitions..

Course Content

  • Introduction –> 1 lecture • 3min.
  • Feature Selection Introduction –> 1 lecture • 2min.
  • Filter Method –> 15 lectures • 2hr 10min.
  • Wrapper methods –> 14 lectures • 54min.
  • Embedded Methods for Feature Selection –> 9 lectures • 32min.

Complete Bootcamp 2021 : Feature selection using Python

Requirements

  • Familiarity with Python programming.
  • Working knowledge of Jupyter Notebook.
  • Working Knowledge of Pandas and Numpy.
  • Working Knowledge of Machine learning Model Creation using sklearn.
  • Understanding of Statistical methods like chisquare test.

Feature selection is one of most important activity in machine learning/Artificial Intelligence pipeline. We select all relevant features for machine learning algorithm and discard less relevant or not relevant features. Feature selection is also known as  variable selection.This course will provide learner, detailed knowledge of feature selection. It is one of most detailed online course on feature selection.

Who is this course for ?

  • Data scientist who wants to create faster and more interpretable machine learning models.
  • Data analyst who wants to relation between two variables.
  • Data science aspirants who are preparing for data science interview.
  • Any One who wants to learn about feature selection process.
  • AI/ML software engineer who write code for machine learning.
  • Teachers who are teaching Machine Learning Models.

What will you learn ?

  • In this course, you are going to learn feature selection by doing. I have included more than 8 end to end small projects on feature selection methods. Each method has one project so that learner can understand the process fully. Code provided in throughout course is downloadable. You can download code and data and run by yourself to get confidence. Knowledge gain though this course is precious and can be used in   We are going to learn following topics.

What is feature selection?

Different methods of feature selection.

Filter methods

  • Minimum variance method
  • F-Score using correlation for regression analysis data.
  • Anova F for classification analysis data
  • Mutual Information for regression and Classification analysis data.
  • Chi-Square Scores for categorical features and Target
  • All these methods implementation using sklearn

Wrapper Method

  • Forward selection of features.
  • Backward selection of features.
  • Exhaustive feature selection.
  • Implementation of each using sklearn and mlxtend.

Embedded Method

  • Introduction to Embedded Method for feature selection.
  • Using RandomForest
  • Using Extremely randomized trees to select features
  • Regularization based feature selection

So what are you waiting for? Join the course and get the knowledge of variable selection and apply it in your projects to get efficient and interpretable machine learning models.