Andrew Ng's Coursera Courses: My Dive into Machine Learning

When ChatGPT launched in late 2022, I got hooked. I have to admit, as someone who spent 18 years in advertising with zero coding background, the world of machine learning felt both fascinating and intimidating. I asked around my network and even asked ChatGPT itself — “where should I start?” — and the answer kept coming back to Andrew Ng's courses on Coursera. So I signed up for the “Machine Learning Specialization” and “Deep Learning Specialization” by Stanford and Deeplearning.AI with Andrew Ng as the instructor.

The part where I had no idea what I was doing

The Machine Learning Specialization was my first foray into the domain. Created by Stanford Online and DeepLearning.AI, this beginner-friendly program promised a comprehensive introduction to machine learning fundamentals, from building ML models with NumPy & scikit-learn to applying unsupervised learning techniques.

When it finally clicked

The program gave me a solid foundation — building ML models using Python, NumPy, and scikit-learn. I learned to construct and train supervised models for prediction and binary classification tasks. Andrew Ng is an incredible teacher. His ability to explain complex concepts simply is honestly unmatched.

The part where I almost quit

The biggest challenge for me was Python and the command line interface. I had never written a line of code before :D But I had ChatGPT by my side, and I just kept asking questions — even the really basic ones. Slowly but surely, things started making sense. I think the key for me was not being embarrassed to ask "dumb" questions.

Going deeper (and getting humbled again)

Next on my learning agenda was the Deep Learning Specialization. This program, also instructed by Andrew Ng, dives deeper into the intricacies of machine learning, focusing on neural network architectures and cutting-edge techniques.

Where it got steep

Going from the basics to deep learning was steep, I won’t lie. However, the curriculum is well-structured and the hands-on projects really helped. There is a bit of overlap between this course and the one above. Since it is online learning, you can fast forward or skip the content you have already covered.

When theory met reality

The real-world projects were the best parts. They bridge the gap between theory and practice, and I could start to see how these concepts could be applied to actual problems — which is what got me excited about building my own chatbot later on.

The moment I started building things

Lastly, the course on Building Systems with the ChatGPT API opened up a new avenue for me to explore the integration of large language models into practical applications. Based on what I have learned from the previous two courses, I managed to:

• Export data from my website, which is built on Wordpress
• Clean and prepare the data. You can check out some sample code here. Of course, the code needs to be revised further based on the project /API specification.
• Use embedding API to build a better search function using LLM.

Making different pieces talk to each other

This course taught how to automate complex workflows using chain calls to a large language model — basically how to get different parts of your application to talk to each other through the LLM.

Looking back at the whole journey

Looking back, the journey was tough but incredibly rewarding. Each course expanded not just my understanding, but my sense of what is possible. From someone who didn't know what a command line was to someone who could build things with code — that shift still feels surreal to me.

And guess what? I succeeded in building my chatbot using OpenAI API, embedding technology, etc... You can check out the lessons learned and the chatbot here.

Have you taken any of Andrew Ng's courses? Or are you thinking about it? I'd love to hear about your experience — especially if you are also coming from a non-technical background like me.

Cheers,

Chandler