Machine learning has revolutionized the world in a very short span of time. Since the data is growing at an exponential rate we need to learn how to process it and gain insights from it.
Machine learning is a field that combines stats and software development into one profession whose primary goal is to build predictive models.
In order to stay top of the field, we need to constantly learn new things. One of my favorite ways is to learn from smarter people and, if possible, do that for free.
And one of the most effective ways to do so is to subscribe to the best machine learning YouTube channels. It’s a great source of knowledge, the latest trends, and an easy way to develop new skills.
In this article, we will go through the top 14 YouTube channels for you to harness yourself with the knowledge of machine learning.
If you are someone who likes to understand everything from scratch then this is by the far the best YouTube channel to learn about Machine Learning.
If you are keen to learn every algorithm’s workflow like how does bias and intercept get updated at every epoch, or how to implement a given machine learning algorithm from scratch then you must check the following series made by Harrison Kinsley himself.
2. Data School
Kevin Markham who is the founder of dataschool.io and the owner of the YouTube channel Data School educates machine learning enthusiasts. You can get a comprehensive understanding of machine learning regardless of your educational background thanks to Kevin‘s teaching.
Kevin also makes videos that cover several tools like pandas, NumPy, scikit-learn that will help you build your machine learning models.
You can binge-watch the following series created by Kevin to get a good grasp of the machine learning fundamentals.
Channel: Data School
3. Artificial Intelligence – All in One
The courses on Artificial Intelligence – All in One covers topics like text mining, text retrieval, and search engines, Neural Networks, and Computer Vision.
You may want to check out the following series to get an excellent grasp over machine learning concepts which is taught by Andrew Ng himself.
“Deep Learning is a superpower. With it, you can make a computer see, synthesize novel art, translate languages, render a medical diagnosis, or build pieces of a car that can drive itself. If that isn’t a superpower, I don’t know what is.” – Andrew Ng
If you want to dive deep into deep learning then you should check out the following series.
5. Machine Learning with Phil
Phil Tabor is a machine learning engineer who creates educational videos in the domain of machine learning and deep learning.
He has created a great playlist regarding Deep Reinforcement Learning tutorials where he is teaching the core concepts of reinforcement learning like deep deterministic policy gradients in TensorFlow 2, Soft actor-critic in PyTorch, Robotic Control with TD3, and many more.
Channel: Machine Learning with Phil
6. Jeremy Howard
Jeremy Howard is a data scientist who has an educational background in philosophy but later out of the curiosity he harnessed himself with the knowledge of stats and programming to build the most effective and easy-to-use library for deep learning tasks fastai.
Making deep learning models was never that easy before fast.ai came into the picture. If you are someone who wants to build deep learning models that complete the task in the field of computer vision like image segmentation, image classification, image restoration with the minimal coding, and maximum results then the fast.ai is suitable for you.
You might wanna binge watch the following series to get a good grasp of deep learning with the help of the fast.ai library.
Channel: Jeremy Howard
7.Two Minute Papers
Two Minute Papers is an awesome channel for anyone who loves to be updated with the latest research going on in the Machine Learning domain.
Two Minute Papers make 2 minutes (almost) long videos explaining a research paper.
If you are keen into research field then you may want to check out the following series
Channel: Two Minute Papers
8. Lex Fridman Podcast
Lex Fridman Podcast is one of the most popular and best machine learning YouTube channels. Its host is an AI researcher working on autonomous vehicles, human-robot interaction, and machine learning at MIT and beyond.
Lex talks with his guests on everything related to AI and ML. But he’s not limiting himself only to this theme. He talks about other things that can inspire, teach, and push you to exceed your limitations.
Insights from all the superstars, influencers, and leading scientists from the world of machine learning. He interviewed such personas like Elon Musk, Nick Bostrom, Andrew Ng, Yann LeCun, Vladimir Vapnik, Mat,t Botvinick and many, many more.
He also has a second YouTube channel called Lex Clips where he posts clips from his podcast and other videos.
Kaggle channel is a spot on YouTube where you can dive into the world of Kaggle community, learn, and do your data science work.
The channel offers videos with interviews with data scientists, lessons, and insightful tips.
This is one of the best machine learning YouTube channels for everyone who wants to learn tricks, experiment, and implement new practices into their own work, no matter what environment you work in.
10. Arxiv Insights
Arxiv Insights is a channel owned by Xander Steenbrugge. He summarizes his core takeaways from a technical point of view while making them accessible for a bigger audience.
If you love technical breakdowns on ML and AI but want a nice summary of the difficult and technical topics, that’s the right place for you!
Although the author doesn’t upload videos often on a regular basis, the channel is praised for its interesting content.
11. Google Cloud Platform
It would be a sin not to subscribe to Google Cloud Platform. On the channel, you can get to know such topics as secure infrastructure, developer tools, APIs, data analytics, and machine learning.
This ML YouTube channel lets you learn about how things work at Google, how to become a better data scientist, and all things Google.
Here’s an interesting video about Google Data Center Security: 6 Layers Deep
DeepLearning.TV is all about Deep Learning. The channel features topics such as How To’s, reviews of software libraries and applications, and interviews with key individuals in the field.
There’s also a series of concept videos showcasing the intuition behind every Deep Learning method so you can better understand how deep learning works.
Springboard channel is all about data science. There are data science and machine learning talks with experts from the leading companies, Women in Data Science playlist with interesting conversations with women who work in ML, deep dives, or mini lessons.
It’s a great machine learning YouTube for those who want to learn how to get a job, what to pay attention to, and find out what it means to work in data science.
14. The TWIML AI Podcast with Sam Charrington
If you’re looking for the latest news from the world of machine learning, make sure to check out the TWIML (This Week in Machine Learning) Podcast YouTube channel.
Here, you’ll find each week’s most interesting and important stories from the world of machine learning and artificial intelligence. It’s a great source of information and knowledge for everyone who wants to stay on top of the latest trends, innovations, and get interesting insights from the experts of ML.
We have listed out the best channels to learn machine learning for free, but you have to be adamant to learn it and you can only learn machine learning by putting your knowledge into practice.
Best of luck on your machine learning journey. 🙂
ML Experiment Tracking: What It Is, Why It Matters, and How to Implement It
Jakub Czakon | Posted November 26, 2020
Let me share a story that I’ve heard too many times.
”… We were developing an ML model with my team, we ran a lot of experiments and got promising results…
…unfortunately, we couldn’t tell exactly what performed best because we forgot to save some model parameters and dataset versions…
…after a few weeks, we weren’t even sure what we have actually tried and we needed to re-run pretty much everything”
– unfortunate ML researcher.
And the truth is, when you develop ML models you will run a lot of experiments.
Those experiments may:
- use different models and model hyperparameters
- use different training or evaluation data,
- run different code (including this small change that you wanted to test quickly)
- run the same code in a different environment (not knowing which PyTorch or Tensorflow version was installed)
And as a result, they can produce completely different evaluation metrics.
Keeping track of all that information can very quickly become really hard. Especially if you want to organize and compare those experiments and feel confident that you know which setup produced the best result.
This is where ML experiment tracking comes in.Continue reading ->