- 26
- 573 587
WelcomeAIOverlords
United States
Приєднався 23 чер 2019
My name is Zak Jost and I'm currently working as a Research Scientist at Amazon Web Services in the Seattle area. The point of this channel is to help people like you level up their Machine Learning skills.
The current focus of this channel will be on Video lectures and Hands-on Screencasts. For lecture material, the focus is on stripping away unnecessary formality and shining a light on the key points to build your intuition. However, I am a firm believer that we learn best by doing, so the screencasts will compliment the lectures by walking you through building things from scratch.
Outside of UA-cam, I also provide long-form blogs as a textual compliment. To keep up with the content announcements across all outlets, you can subscribe to my Email List using the link below.
The current focus of this channel will be on Video lectures and Hands-on Screencasts. For lecture material, the focus is on stripping away unnecessary formality and shining a light on the key points to build your intuition. However, I am a firm believer that we learn best by doing, so the screencasts will compliment the lectures by walking you through building things from scratch.
Outside of UA-cam, I also provide long-form blogs as a textual compliment. To keep up with the content announcements across all outlets, you can subscribe to my Email List using the link below.
How to transform tabular data into a graph (walk-thru)
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)!
Code for this project: github.com/zjost/blog_code/tree/master/tab2graph_fraud
Discord Server: discord.gg/xh2chKX
Chapters
- 00:00 - Running the notebook
- 00:56 - Overview of content
- 01:53 - Relational databases as graphs
- 04:00 - Tabular datasets as graphs
- 08:33 - Q: Better to have more nodes or more relationships?
- 09:15 - Q: Can we use multiple edge types?
- 09:30 - Q: How do you build similarity-based connections?
- 11:24 - Q: How to decide between representing a column as a node or node property?
- 13:00 - Configuring the notebook environment
- 14:29 - Looking at example fraud data to design the graph
- 22:33 - Creating the nodes
- 25:00 - Creating the edges
- 26:50 - Instantiating the graph
- 27:32 - Attaching labels to the nodes
- 30:00 - Visualizing the graph
- 35:07 - Constructing node features
- 36:22 - Baseline (non-graph) models
- 41:40 - Label propagation
- 44:19 - Graph neural network
- 46:51 - Handling featureless nodes
- 51:21 - Q: How to figure out if test nodes are connected to train nodes?
- 53:33 - Q: How to split into train/test
- 54:21 - An alternative graph structure and time-directed edges
- 57:07 - Q: On the interaction of time and graph structure
- 58:45 - Q: Should I remove strong features to troubleshoot the model?
Code for this project: github.com/zjost/blog_code/tree/master/tab2graph_fraud
Discord Server: discord.gg/xh2chKX
Chapters
- 00:00 - Running the notebook
- 00:56 - Overview of content
- 01:53 - Relational databases as graphs
- 04:00 - Tabular datasets as graphs
- 08:33 - Q: Better to have more nodes or more relationships?
- 09:15 - Q: Can we use multiple edge types?
- 09:30 - Q: How do you build similarity-based connections?
- 11:24 - Q: How to decide between representing a column as a node or node property?
- 13:00 - Configuring the notebook environment
- 14:29 - Looking at example fraud data to design the graph
- 22:33 - Creating the nodes
- 25:00 - Creating the edges
- 26:50 - Instantiating the graph
- 27:32 - Attaching labels to the nodes
- 30:00 - Visualizing the graph
- 35:07 - Constructing node features
- 36:22 - Baseline (non-graph) models
- 41:40 - Label propagation
- 44:19 - Graph neural network
- 46:51 - Handling featureless nodes
- 51:21 - Q: How to figure out if test nodes are connected to train nodes?
- 53:33 - Q: How to split into train/test
- 54:21 - An alternative graph structure and time-directed edges
- 57:07 - Q: On the interaction of time and graph structure
- 58:45 - Q: Should I remove strong features to troubleshoot the model?
Переглядів: 15 104
Відео
Graph Attention Networks (GAT) in 5 minutes
Переглядів 36 тис.2 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt) GAT paper: arxiv.org/abs/1710.10903 Mailing List: blog.zakjost.com/subscribe Discord Server: discord.gg/xh2chKX
Author Interview - Equivariant Subgraph Aggregation Networks
Переглядів 1,4 тис.2 роки тому
Paper link: arxiv.org/abs/2110.02910 Primary authors: Beatrice Bevilacqua (@beabevi_), Fabrizio Frasca (@ffabffrasca), Derek Lim (@dereklim_lzh) Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)
PhD unboxing (?) + Channel Update
Переглядів 2,1 тис.2 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt) Mailing List: blog.zakjost.com/subscribe Discord Server: discord.gg/xh2chKX Blog: blog.zakjost.com Patreon: www.patreon.com/welcomeaioverlords
Scaling Graph Neural Networks to Twitter-scale (w/ Emanuele Rossi interview)
Переглядів 3,1 тис.3 роки тому
This is an interview with Emanuele Rossi, who is an ML researcher at Twitter and Phd student with Michael Bronstein that focuses on applying GNNs to large scale, dynamic graphs. Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt) You can find more from Emanuele at: www.emanuelerossi.co.uk/ SIGN paper: arxiv.org/abs/2004.11198 Mailing List: blog...
Neural Networks as Guitar Amps (with Neural DSP interview)
Переглядів 7 тис.3 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! This video talks about how Machine Learning is beginning to have an impact in the world of guitar circuit modeling. It also contains an interview with Neural DSP ML Researcher Lauri Juvela. The discussion thread with Lauri is here: www.welcomeaioverlords.com/t/guitarml-w-lauri-juvela-from-neu...
Graph Convolutional Networks using only NumPy
Переглядів 38 тис.3 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! Implements Graph Convolutional Networks from scratch to translate the paper's equations into code. Applies this to Zachary's Karate Club graph as an example. Code: github.com/zjost/blog_code/tree/master/gcn_numpy Original paper: arxiv.org/pdf/1609.02907.pdf Message Passing video: ua-cam.com/v...
Simple Message Passing on Graphs
Переглядів 30 тис.3 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! This video discusses the adjacency matrix and how it can be used to implement basic message passing on graphs. A simple example is given using Python. Code: github.com/zjost/blog_code/blob/master/gcn_numpy/message_passing.ipynb Mailing List: blog.zakjost.com/subscribe Discord Server: discord....
How DeepMind learns physics simulators with Graph Networks (w/ author interview)
Переглядів 22 тис.3 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! This video dives into the paper "Learning to Simulate Complex Physics with Graph Networks" from DeepMind and interviews one of its authors, Jonathan Godwin. Original Paper: arxiv.org/abs/2002.09405 Simulator video source: sites.google.com/view/learning-to-simulate/ Project Code & Datasets: gi...
Intro to Relational - Graph Convolutional Networks
Переглядів 30 тис.3 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! This video presents Relational Graph Convolutional Networks (R-GCNs) as a way to apply GCNs to heterogeneous graphs. A simple Twitter graph is used to demonstrate the concepts. Both regularization techniques presented in the paper (Basis- and Block Diagonal-Decomposition) are also covered. Pa...
OpenAI API (GPT-3) is magical...
Переглядів 18 тис.4 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! Some capabilities of the new OpenAI API that wraps a GPT-3 model. Demos are from OpenAI announcement of the API. Mailing List: blog.zakjost.com/subscribe Discord Server: discord.gg/xh2chKX Blog: blog.zakjost.com Patreon: www.patreon.com/welcomeaioverlords Attribution: - Original demo videos: ...
Graph Convolutional Networks (GCNs) made simple
Переглядів 115 тис.4 роки тому
Join my FREE course Basics of Graph Neural Networks (www.graphneuralnets.com/p/basics-of-gnns/?src=yt)! This video introduces Graph Convolutional Networks and works through a Content Abuse example. For a hands on example with code, check out this blog: blog.zakjost.com/post/gcn_citeseer/ Mailing List: blog.zakjost.com/subscribe Discord Server: discord.gg/xh2chKX Original Paper: openreview.net/p...
Intro to Graphs and Label Propagation Algorithm in Machine Learning
Переглядів 37 тис.4 роки тому
Intro to Graphs and Label Propagation Algorithm in Machine Learning
How to use Adversarial Validation to Help Fix Overfitting
Переглядів 3,7 тис.4 роки тому
How to use Adversarial Validation to Help Fix Overfitting
Five surprises of joining a tech company
Переглядів 9584 роки тому
Five surprises of joining a tech company
Using Neural Networks for Credit Card Fraud Detection
Переглядів 7 тис.4 роки тому
Using Neural Networks for Credit Card Fraud Detection
Ensembling Machine Learning Models for Fraud Detection
Переглядів 2,2 тис.4 роки тому
Ensembling Machine Learning Models for Fraud Detection
Build Credit Card Fraud Detection ML Model from Scratch
Переглядів 6 тис.4 роки тому
Build Credit Card Fraud Detection ML Model from Scratch
How to Create a new Python ML Project (Screencast)
Переглядів 1,7 тис.4 роки тому
How to Create a new Python ML Project (Screencast)
Understand the Math and Theory of GANs in ~ 10 minutes
Переглядів 61 тис.4 роки тому
Understand the Math and Theory of GANs in ~ 10 minutes
Gentle Intro to Generative Adversarial Networks - Part 1 (GANs)
Переглядів 14 тис.4 роки тому
Gentle Intro to Generative Adversarial Networks - Part 1 (GANs)
Anybody told you you look exactly like Chad Michael Murray from One Tree Hill? The resemblance is remarkable
Awesome 👍
That's very informative and well explained video. Great job and thank you.
Are you and 3 Blue 1 Brown the same person?
No, Grant Sanderson is far more prolific and talented :)
Best mathematical explanation on GAN on the internet so far
Can you send the intro of this video by somehow , I really liked the choice of music and the transitions you made.
The song is "Legend" by the band Elder.
Good to see that you are doing your part to bring Roko's Basilisk into existence with your UA-cam channel.
I don't think any other explanation can do better than this. Thanks a lot!
Best straight-forward explanation i've ever seen. Thanks
very helpful, thank you !
Bro gave me a great intuitive and clear explanation without using fancy pictures. Greatest RGCN explanation ever.
amazingly explained
❤🔥
I preffer cocatenating feature vectures and apply the weighted sum. in the agregation phase. averaging to me is a crime
Error in Min. 2:53 --> alpha u,v expression --> denominator exponent should be a(hk, hv)
is there any article about this?
great channel! thank you for the awesome explanation!
Awesome Video
Great!
One of the best explanation ever
nice
“A neural network can learn any function, including the probability distribution of images of faces.” Woah
what do u mean by the degree of a node?
Awesome explanation man.. I loved it!!
Thank you, Zak. It helps me well.
literally made it so simple. amazing explanation
Clear, honest, patient explanation. Thank you. You're doing this right, and you're a big asset to people who want to learn. Please don't change your approach at all.
brilliant explanations!
Hi, Zak thank you for your great videos!! I would like to ask around, is that possible or easy to translate IP network traffic packets into graph-structured data?
Amazing video!
Zak come back and do more videos on UA-cam
I've been thinking about it. Any ideas for videos?
Either another "from scratch" video, like "relational GCNs from scratch using only Numpy" for example or something on the training side, like how negative sampling can be used effectively in GNN training (I'm thinking about the ItemSage paper specifically as an example)
THE BEST VIDEOS LITERALLY! Awesome!!!!!!
You're awesome! Absolutely amazing explanations!! THANK YOU
Brilliant
Holy sh*t, this guy is diabolically, criminally, offensively underrated. THE best explanation of GANs I have ever seen, somehow rooting it deeply in the mathematics while keeping it surface level enough to fit in a 12 min video. Wow
This whole series is excellent.
Superb
Great Explanation! Nice and clean! All the best
Man you explain things so freaking sharply, in ways no one thinks of - this should have at least 500k views
hey, thanks for the info, just a question, in case of when the node n1 is an item (not person) having some feature-vector v1 connected to multiple neighbours Ni's with feature vector Vi's can R-GCN be prone to target leakage cause it's aggregating the node-embeddings?
Holy shit. Never have I watched a UA-cam video (let alone on a complex topic like this) and had to *pause* so many times because something that never clicked just blew my mind. If everyone could explain things like you, the world would be a better place. Thank you 🙏
Love it!
Amazing video! Thanks so much!
This video is amazing! You make things intuitive and really dig down to the core idea. Thank you! And also subscribed your blog!
Great learning resource for GAN. Thank you.
This is one of the best explanations I have seen! Great video!
That's the most understandable introductions on GAT I've ever seen! Really love the way you explain very complicated things using simple examples. Thank you so much! Any chance you can do a video on transformer some time?
Yo how to graph auto encoders work? Great video
The hallmark of a good teacher is when they can explain complex topics simply and intuitively. And your presentation on GANs in this video truly marks you as a phenomenal one. Thanks!
I agreeeeeeeee, you are the best, thank you sooo mutch
Simply beautiful ... many thanks for this video !