So you’re considering Udacity’s Intro to Machine Learning with PyTorch Nanodegree? I went through and finished this program in 2019, and in this video, I’m going to provide my Udacity Nanodegree review..coming up.

First, I’m going to describe the courses and projects in the Udacity’s Intro to Machine Learning with PyTorch Nanodegree. Next, I’ll talk about factors such as the extracurricular courses, lecture videos, and project reviews in the Nanodegree. And finally, I’ll give my honest opinion on the Udacity’s Intro to Machine Learning with PyTorch Nanodegree, overall.

There are three courses in this nanodegree program.

Courses & Projects

Course 1: Supervised Learning

In this first course, you will learn about supervised learning, which is a common class of methods for machine learning model construction.

Project: Find Donors for CharityML

The project associated with this course is called “Find Donors for CharityML.” It involves writing a Python script using a dataset from a fictional charity organization. The purpose of the project is to identify categories of people (based on existing charity donors in the dataset) that are most likely to donate to the charity. You’ll need to evaluate and optimize at least three different supervised learning algorithms to determine which algorithm will provide the highest donation yield.

Examples of supervised learning models you might use are:

• Logistic Regression
• Naïve Bayes
• Support Vector Machine (SVM)
• Decision Trees
• Random Forest

This project was relatively straight-forward and if you have at least basic experience with Python, you shouldn’t have issues with completing it. By the end of this project, your knowledge of supervised learning algorithms, model evaluation methods, and feature importances will definitely increase.

Course 2: Neural Networks

In the second course, you’ll  learn the foundations of neural network design and training in PyTorch.

Project: Build an Image Classifier

The project associated with this course is called “Build an Image Classifier.” In this project you’ll implement an image classification application using a deep neural network in PyTorch. This image classification application will train a deep learning model on a dataset of images. It will then use the trained model to classify new images.

This project is consdierably more difficult than the first project. I spent more time on this deep learning project than the other two projects combined in this Nanodegree program. I recommend at least an intermediate skill level in Python before attempting this project. In particular, you’ll want to be well-practiced with Python functions, dictionaries, and lists. Having some initial knowledge of the Python Imaging Library (PIL) might prove beneficial as well.

Course 3: Unsupervised Learning

In the third course, you’ll learn to implement unsupervised learning methods for different kinds of problem domains.

Project: Create Customer Segments

The project associated with this course is called “Create Customer Segments.” In this project, you will utilize clustering algorithms such as K-means and Principle Component Analysis to compare a business’s customer data to external demographic data to identify over and under-represented customer populations.

A very high percentage of this project involves pre-processing the data and creating plots with either the Seaborn package in Python or the MatPlotLib package in Python. So I would recommend going into the project with at least a basic skill level in those two visualization packages.

Other Considerations

When it comes to the lecture videos, they’re among the best I’ve seen and they’re very informative. The presentations from the speakers and the use of animations were incredibly helpful, especially in the sections about supervised learning and deep learning.

When it comes to the project reviews, I typically received a review within a few hours after submitting a project. If there were issues with a project I submitted, those issues were clearly communicated by my reviewer.

The mentor support is helpful when it comes to challenges that are relatively common in the projects.

And I found the resume and LinkedIn profile review services valuable as well, which are available through their Career Resource Center.

Final Thoughts

Overall, it was a great learning experience and I continued to use many of the Machine Learning and Python concepts from this program in work situations afterwards. But, deep learning isn’t as widely used in many organizations as machine learning at this point of time and that should be taken into consideration.

To cap off, I firmly believe that project-based learning is the best way to learn Machine Learning, Deep Learning, and Python and in that regard, this program is one of the best educational options available.

___________________

URLs:

Udacity “Intro to Machine Learning with PyTorch” Nanodegree
🎓 https://bit.ly/35Wp6cL

Resources to Help you Prepare for the Udacity Data Science Nanodegree:

Udacity “Programming for Data Science with Python” Nanodegree:
🎓 https://bit.ly/2O7mwqj

Udacity Data Analyst Nanodegree:
🎓 https://bit.ly/3cidu55

Official PyTorch Tutorials:
💻 https://pytorch.org/tutorials

Python Data Science Handbook: Essential Tools for Working with Data:
📚 https://amzn.to/36xsYBi