Numerical Methods  of Machine Learning

Modules Overview

9 Modules

Module 1. Introduction

Start by getting familiar with the structure of the course, key concepts, tasks, and the final project. You'll understand how numerical methods support model training and how this course connects mathematical techniques to machine learning applications.

Module 2. Entry Assessment

1.5h

Assess your current knowledge of numerical methods in machine learning. The test is divided into topic-specific blocks and will help you identify your strengths and the areas that need improvement. After completing the test, you'll receive instant feedback and personalized learning recommendations.

Module 3. Algorithm Analysis

Understand why numerical methods matter by analyzing the complexity and structure of training algorithms. You’ll learn to calculate computational complexity, distinguish between direct and iterative methods, and solve equations using the bisection method.

Module 4. Gradient Descent

Learn how gradient descent works — one of the foundational methods behind modern training algorithms. You’ll study loss functions in classification and regression tasks, explore how gradients are calculated, and break down the gradient descent algorithm.

Module 5. Gradient Descent Training

1.5h

See how gradient descent is used to train linear regression models and neural networks. You’ll write an algorithm for stochastic gradient descent, understand regularization techniques, and explore how neural networks learn.

Module 6. Gradient Boosting

Take your training algorithms to the next level by exploring gradient boosting. You’ll learn how this method builds powerful predictive models by combining weak learners, how to reduce overfitting, and how to work with the CatBoost library.

Module 7. Course Project

Put your skills to the test with a practical business challenge. You’ll work with Rusty Bargain, a used car sales platform, to build a model that predicts car values. Using historical sales data, technical specs, and trim features, your model will need to optimize for prediction quality, speed, and training time.

You’ll submit your project for expert review, receive feedback, and make improvements until it meets quality standards.

Module 8. Final Assessment

1.5h

Evaluate your learning progress with a final skill assessment. Compare your results to the entry test and see how far you've come. Your test results will be broken down by topic, with immediate feedback to guide your next steps.

Module 9. Conclusion

Wrap up the course by reviewing key takeaways and leaving feedback. You'll finish with a solid understanding of how numerical methods power real machine learning solutions.

Numerical Methods of Machine Learning

Key Program Takeaways

Modules Overview

Get in Touch!

Explore more ML courses