Data Prep for Machine Learning in Python

In this course, we'll learn how to import and clean data before populating missing values using imputation. We'll learn how to visualize histograms, scatter charts, and box plots to identify trends of interest before using the analysis to select the most important features. Feature engineering techniques such as one hot encoding, binning and scaling will help us transform the structure of our data to produce higher quality machine learning insights.

Skills you'll gain

This course includes:

This module covers essential techniques for importing data from various sources like CSV, Excel, and SQL, and performing initial data cleaning operations. Topics include selecting columns, filtering rows, data type validation, comparing datasets to trusted sources, and various imputation methods for handling missing data. The module includes practical exercises on importing and filtering data, data validation, and cleaning and imputation techniques to prepare data for further analysis.

Skills you'll gain

This course includes:

This module focuses on exploring and understanding data through statistical analysis and visualization techniques. It covers descriptive statistics for numeric features, basic plots for both numeric and categorical features, and methods for analyzing relationships between variables. Topics include univariate analysis with histograms and box plots, multivariate analysis examining continuous-continuous and categorical-continuous variable relationships, and creating and interpreting various visualization types to identify patterns in the data.

Skills you'll gain

This course includes:

This brief module reviews the concept of train-test splits in machine learning. It explains the importance of separating data into training and testing sets, and demonstrates how to implement train-test splits using the scikit-learn library in Python. The module serves as a refresher on this fundamental concept before moving on to more advanced data preparation techniques.

Skills you'll gain

This course includes:

This assessment module tests students' understanding and application of the concepts covered in the first part of the course, including data importing, cleaning, exploratory data analysis, and train-test splitting. Students apply these techniques to solve practical data preparation problems.

Skills you'll gain

This course includes:

This module covers essential feature engineering techniques focusing on encoding categorical variables and transforming feature distributions. It begins with reviewing train-test splits before diving into encoding theory, with emphasis on one-hot encoding. The module demonstrates practical implementation of one-hot encoding using both pandas and scikit-learn, comparing their differences. It then explores distribution transformation techniques for handling skewed data, including identifying skew in Python and applying appropriate transformations such as logarithmic transformation. The module includes exercises to reinforce these concepts through practical application.

Skills you'll gain

This course includes:

This comprehensive module explores advanced feature engineering techniques. It starts with outlier management, covering both removal and modification approaches. The module then delves into various binning strategies including categorical binning, width-based binning, frequency-based binning, and manual binning, with a discussion on smoothing techniques. The final section covers feature scaling, explaining why it matters and demonstrating different scaling methods like min-max scaling and standard scaling, with emphasis on properly scaling test data. Throughout the module, exercises help students apply these techniques to real data, and it concludes with guidance on making appropriate feature engineering decisions based on specific modeling contexts.

Skills you'll gain

This course includes:

This module focuses on techniques for identifying and selecting the most relevant features for machine learning models. It covers manual feature selection as well as statistical methods based on the target variable type. For continuous target variables, it demonstrates correlation coefficients for continuous features and ANOVA for categorical features. For categorical target variables, it explains how to use box plots for continuous features and chi-square tests for categorical features. The module provides a comprehensive overview of different feature selection approaches and when to apply each technique.

Skills you'll gain

This course includes:

This brief module summarizes the key concepts covered throughout the Data Prep for Machine Learning in Python course. It recaps essential techniques for data importing, cleaning, exploratory data analysis, feature engineering, and feature selection, providing a comprehensive conclusion to the course material.

Skills you'll gain

This course includes:

This final assessment module tests the students' understanding of feature engineering and feature selection techniques covered in the course. Students apply their knowledge to solve complex data preparation problems, implement various feature engineering methods, select appropriate features, and prepare data for machine learning applications.

Skills you'll gain

This course includes: