TL;DR: What's your budget? I'm pretty cautious about pricey stuff, but I'm basically super satisfied with Brilliant.org App. It's interactive and actually affordable compared to some... what's your limit?? lol

Saw this earlier but just now responding! Honestly, for ur commute, these are pretty solid:

* Mimo: Learn to Code and AI App: Has great Python/ML paths. Very hands-on with a mobile IDE, but its a bit simplified for advanced stuff.
* Enki: Learn Data Science: Uses spaced repetition for ML theory and neural networks. Awesome for retention, tho coding challenges are limited.

Id go with Mimo cuz typing the code actually makes it stick. gl!

> I'm diving into machine learning and want to make the most of my commute! I’m looking for a top-rated app that covers everything from basic regression to neural networks. I'd love something with interactive quizzes or hands-on coding challenges rather than just videos.

hey! honestly i've been in the ML space for like 8 years now and I've tried many different tools to keep my skills sharp during downtime... In my experience, I would suggest DataCamp: Learn Data Science. It's basically the industry standard for interactive learning. Over the years, i've found their mobile interface is actually reallyyy good for hands-on coding challenges cuz they use a custom UI so you dont have to type every bracket on your phone. It covers everything from basic regression to deep learning.

If you want the theory to really click, I also recommend Brilliant.org Premium Subscription. It's less about coding and more about the logic/math, which makes things like neural networks actually stick. tbh, most apps are just fluff but these two are legit. anyway, hope that helps! gl!

^ This. Also, I’ve been following this thread for an hour and honestly it drives me crazy how much these ed-tech companies are charging these days. I had high hopes for using my commute to actually level up but unfortunately my experience has been a total letdown. Its like they care more about flashy animations than actually teaching the hard stuff.

• the subscription prices are honestly getting ridiculous for the tiny bit of value you get back.
• most of these apps feel like they’re just chasing the hype and dont actually update their core curriculum anymore.
• its such a scam when you pay for a premium tier and the hands-on part is just a broken console that crashes on your phone.
• it feels like the quality is just going downhill while they keep asking for more money every month. Tbh it just makes me feel so defeated. I spent so much time and cash on these platforms and still feel like I'm hitting a wall... it isnt just you, the whole state of these learning apps right now is just frustrating as hell. If youre feeling burnt out by the lack of depth, I totally get it.

> I'd love something with interactive quizzes or hands-on coding challenges rather than just videos. Been thinking about your question for a bit and honestly, my journey taught me to be extremely cautious with the gamification of ML. While those bite-sized quizzes feel productive during a commute, I found that they often lead to a false sense of mastery because they basically abstract away the actual complexity of the underlying mathematics. In my professional experience, the biggest mistake is choosing a platform that ignores the rigorous side of things like:

• Linear algebra fundamentals
• Proper bias-variance tradeoff analysis
• The actual implementation of loss functions from scratch I eventually shifted away from dedicated learning apps and went the DIY route by using a mobile browser to access cloud-hosted notebook environments. It is definitely more challenging to set up on a phone, but it prevents you from becoming dependent on a simplified UI that does not exist in professional settings. If you rely too much on fill-in-the-blank coding, you might struggle when you have to handle raw data preprocessing or debugging a complex neural network architecture without hints. Just be careful not to trade technical depth for a high score in an app.

> I'm looking for a top-rated app that covers everything from basic regression to neural networks. I'd love something with interactive quizzes or hands-on coding challenges rather than just videos. Honestly, having tracked the evolution of the ML ed-tech market for years, most apps fail because they prioritize flashy UI over actual technical depth. If you want something that actually prepares you for real-world hyperparameter tuning and model evaluation, you should look at Sololearn: Learn to Code. Their ML path is surprisingly robust compared to the more 'casual' competitors, focusing heavily on the Scikit-Learn and Pandas ecosystems which are basically the industry standards for data prep and classical ML. Another solid shout is Udacity. While people usually think of them for long-form nanodegrees, their mobile interface has evolved into a realy efficient way to digest complex concepts like stochastic gradient descent or CNN architectures through interactive "knowledge bytes." From a market analysis perspective, these two bridge the gap between the overly simplified gamification you see in some apps and the high-level theoretical stuff. Definitly helps make those core concepts stick during a commute without feeling like you're just playing a game, lol.

> I'd love something with interactive quizzes or hands-on coding challenges rather than just videos. ^ This. Also, we've seen a variety of suggestions so far, ranging from the big subscription platforms to more casual coding apps, plus a really solid point about not letting gamification replace actual deep learning. Coming back to this, I'd suggest being careful with how much you rely on these mobile tools. They're great for the commute, but make sure to check if you can actually apply the logic without the app holding your hand. From a DIY perspective, I've always found that the more you have to struggle with the code, the better it sticks. A couple things tho... what's your current Python level? And are you looking for something that dives deep into the math or just the high-level stuff?

Just catching up on this thread. Regarding what #6 said about "> I'd love something with interactive quizzes or...", I think the caution about mobile tools is spot on. Most of the apps mentioned so far, like Mimo or SoloLearn, are great for building a habit, but they often lack the technical rigor you'll need for long-term ownership of these skills. If you want something reliable that wont just feel like a game, here are a few thoughts based on what I use to keep my knowledge stable:

• Look into Coursera: Learn new skills. It allows you to download lectures and quizzes for your commute. Its more academic and less flashy, but the content comes from actual universities, which is safer for long-term learning.
• Consider Udacity
• Lifelong Learning. Their ML courses are industry-vetted. While the app is more for viewing, the curriculum is designed to be professional-grade.
• Stick to apps that let you export your progress or offer certificates that actually mean something in the industry. Gamified badges are basically useless once you close the app. Basically, dont get distracted by the UI. Focus on the depth of the documentation provided within the app. If it doesnt explain the why behind the regression math, skip it.

honestly i spent way too long trying to learn on some of the bigger platforms and it was just a letdown. i thought i was making progress but unfortunately the second i tried to optimize a real model i was totally lost. those apps just dont prepare you for when performance starts tanking in the real world... kinda felt like i wasted my commute time for a bit there. i eventually started using Programming Hub AI and ML Path and it was a bit better for actually seeing how the logic flows. it wasnt as good as expected compared to a full desktop setup, but for a phone, it gets the job done. quick tips:

• jump into the compiler as often as possible to see how changing variables impacts the output.
• focus on learning the math behind the error rates instead of just clicking next. i found that helped the concepts stick way more than the quizzes.

Nice, didn't know that