avatar

Proclus Academy

Machine Learning / Deep Learning

Accuracy and Confusion Matrix Using Scikit-Learn & Seaborn

Let's learn how to calculate Confusion Matrix and Accuracy using Python libraries.

I'll also show you two different ways to visualize the Confusion Matrix.

Summary Image: Confusion Matrix and Accuracy Using Scikit-Learn & Seaborn

Using Confusion Matrix and Accuracy to Test Classification Models

Let's look at the basic metrics to estimate a classification model’s predictive performance.

You'll also gain practical skills to generate and visualize these metrics using Scikit-Learn and Seaborn.

Summary Image - Using Confusion Matrix and Accuracy to Test Classification Models

K-Fold Cross-Validation Using Python and Scikit-Learn

How can you perform K-Fold Cross-Validation to evaluate machine learning models?

I'll show you two ways using Python and Scikit-Learn's helper functions - cross_val_score() and cross_validate().

Summary Image: K-Fold Cross-Validation Using Python and Scikit-Learn

What Is K-Fold Cross-Validation?

It's time to learn Cross-Validation, the tool serious data scientists use to estimate model performance.

Cross-Validation builds upon Train Test Split and provides a better estimate of a machine learning model's performance.

Summary Image: K-Fold Cross-Validation

Boxplot With Separate Y-Axis for Each Column

Sometimes you want to draw boxplots where each column gets its own y-axis.

Here's how you can do it using pandas, Matplotlib, and Seaborn.

Boxplot with separate Y-Axis For Each Column: Summary image - Taipei skyline with skyscrapers of different heights

Robust Scaling: Why and How to Use It to Handle Outliers

Outliers can overshadow other data points of a feature. That can negatively influence standard scaling.

Here's why you should use robust scaling to handle outliers.

Robust Scaling Summary Image: A tiny boat and a tall lighthouse.

Summary Statistics Using Pandas

Sometimes the default statistics provided by pandas describe() method are not enough.

Here's how you can generate custom statistics using the agg() method.

Image: Track and field with numbered tracks

Use Standard and MinMax Scaling to Tame Numerical Features

Features with vastly different scales can lead to subpar machine learning models.

We must scale such features. Here’s how sklearn's Standard and MinMax scalers can help.

Image: Half Dome at Yosemite

Use ‘Train Test Split’ to Beat Overfitting

Overfitted machine learning models can lead to serious performance issues in the real world.

Let's explore how you can stop overfitting dead in it's tracks and train your models with confidence.

Featured Image

ColumnTransformer: Why and How to Use It to Preprocess Data

How can you apply different transformations to different columns of your dataset?

We'll explore a few ways and discuss why using ColumnTransformer is the best approach.

Featured Image