An Interview With D. Sculley, CEO Of Kaggle
Skills:
ML Maths Basics60%
This is a clip from an interview between D. Sculley and Lukas Biewald.
๐ฅ Watch The Full Episode At https://www.youtube.com/watch?v=1aajTQvZJ94
D. Sculley is CEO of Kaggle, and author if the influential 2015 paper, " "Machine Learning: The High Interest Credit Card of Technical Debt".
In this interview, he talks with Weights & Biases CEO and Co-founder Lukas Biewald.
๐ Show Notes: http://wandb.me/d-sculley
The transcript of this short of the full interview with D. Sculley:
Lukas Biewald: You are the author of the machine learning "High Interest Technical Debt" paper, which I think inspired a lot of people and really resonated, when it came out, with me.
And, so, I thought maybe you could start โ for people who haven't read this paper โ by kind of summarizing it.
D. Sculley: There are really good reasons to move fast; it's sometimes unavoidable. But in doing so, we create some costs for ourselves over time that need to be paid down. It's not that we can never take those costs on, but we better be honest with ourselves about what those costs are. And, at the time, I think it was under-appreciated how much technical debt can be accrued through the use of machine learning.
Even the simplest things you can think of...like, when you're first building a model, oftentimes if you're in a hurry you rush and you put a whole bunch of features in the model, everything you can think of, you put it in there. Accuracy is .9. You're like, "Okay, that's pretty good, but I can think of another 20 features", and you put all those 20 new features in, and now it's .92. And then you're like, "Well, that's pretty good, but if I put another 20 features in, then I get .93."
So, we made accuracy go up. What could be the problem, right?
As I'm sure you've seen, every time you add a feature into a model, you create a dependency. You now have a dependency on some behavior or observation in the outside world. And this means that you have a vulnerability if that behavior in the outside
Watch on YouTube โ
(saves to browser)
Sign in to unlock AI tutor explanation ยท โก30
Playlist
Uploads from Weights & Biases ยท Weights & Biases ยท 0 of 60
โ Previous
Next โ
1
2
3
4
5
6
7
8
9
10
![6. Recurrent Neural Networks [RNNs]](https://i.ytimg.com/vi/8lbGjKhrJOo/mqdefault.jpg)
11
12
13
![9. Hybrid LSTMs [Long Short-Term Memory]](https://i.ytimg.com/vi/NysY9FN9Uac/mqdefault.jpg)
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0. What is machine learning?
Weights & Biases
1. Build Your First Machine Learning Model
Weights & Biases
Intro to ML: Course Overview
Weights & Biases
2. Multi-Layer Perceptrons
Weights & Biases
3. Convolutional Neural Networks
Weights & Biases
Weights & Biases at OpenAI
Weights & Biases
Why Experiment Tracking is Crucial to OpenAI
Weights & Biases
4. Autoencoders
Weights & Biases
5. Sentiment Analysis
Weights & Biases
6. Recurrent Neural Networks [RNNs]
Weights & Biases
7. Text Generation using LSTMs and GRUs
Weights & Biases
8. Text Classification Using Convolutional Neural Networks
Weights & Biases
9. Hybrid LSTMs [Long Short-Term Memory]
Weights & Biases
Toyota Research Institute on Experiment Tracking with Weights & Biases
Weights & Biases
Weights and Biases - Developer Tools for Deep Learning
Weights & Biases
Introducing Weights & Biases
Weights & Biases
10. Seq2Seq Models
Weights & Biases
11. Transfer Learning for Domain-Specific Image Classification with Small Datasets
Weights & Biases
12. One-shot learning for teaching neural networks to classify objects never seen before
Weights & Biases
13. Speech Recognition with Convolutional Neural Networks in Keras/TensorFlow
Weights & Biases
14. Data Augmentation | Keras
Weights & Biases
15. Batch Size and Learning Rate in CNNs
Weights & Biases
Applied Deep Learning Fellowship Overview and Project Selection with Josh Tobin (2019)
Weights & Biases
Grading Rubric for AI Applications with Sergey Karayev (2019)
Weights & Biases
16. Video Frame Prediction using CNNs and LSTMs (2019)
Weights & Biases
Image to LaTeX - Applied Deep Learning Fellowship (2019)
Weights & Biases
17. Build and Deploy an Emotion Classifier (2019)
Weights & Biases
Applied Deep Learning - Data Management with Josh Tobin (2019)
Weights & Biases
Snorkel: Programming Training Data with Paroma Varma of Stanford University (2019)
Weights & Biases
Applied Deep Learning - Troubleshooting and Debugging with Josh Tobin (2019)
Weights & Biases
Troubleshooting and Iterating ML Models with Lee Redden (2019)
Weights & Biases
Designing a Machine Learning Project with Neal Khosla (2019)
Weights & Biases
Lukas Beiwald on ML Tools and Experiment Management (2019)
Weights & Biases
Building Machine Learning Teams with Josh Tobin (2019)
Weights & Biases
Pieter Abeel on Potential Deep Learning Research Directions (2019)
Weights & Biases
Testing and Deployment of Deep Learning Models with Josh Tobin (2019)
Weights & Biases
Five Lessons for Team-Oriented Research with Peter Welder (2019)
Weights & Biases
Applied Deep Learning - Rosanne Liu on AI Research (2019)
Weights & Biases
Making the Mid-career Leap from Urban Design to Deep Learning/Data Science
Weights & Biases
Organizing ML projects โ W&B walkthrough (2020)
Weights & Biases
Brandon Rohrer โ Machine Learning in Production for Robots
Weights & Biases
Nicolas Koumchatzky โ Machine Learning in Production for Self-Driving Cars
Weights & Biases
My experiments with Reinforcement Learning with Jariullah Safi
Weights & Biases
Applications of Machine Learning to COVID-19 Research with Isaac Godfried
Weights & Biases
Testing Machine Learning Models with Eric Schles
Weights & Biases
How Linear Algebra is not like Algebra with Charles Frye
Weights & Biases
Predicting Protein Structures using Deep Learning with Jonathan King
Weights & Biases
Rachael Tatman โ Conversational AI and Linguistics
Weights & Biases
Reformer by Han Lee
Weights & Biases
Sequence Models with Pujaa Rajan
Weights & Biases
GitHub Actions & Machine Learning Workflows with Hamel Husain
Weights & Biases
Look Mom, No Indices! Vector Calculus with the Frรฉchet Derivative by Charles Frye
Weights & Biases
Jack Clark โ Building Trustworthy AI Systems
Weights & Biases
Surprising Utility of Surprise: Why ML Uses Negative Log Probabilities - Charles Frye
Weights & Biases
Track your machine learning experiments locally, with W&B Local - Chris Van Pelt
Weights & Biases
Antipatterns in open source research code with Jariullah Safi
Weights & Biases
Attention for time series forecasting & COVID predictions - Isaac Godfried
Weights & Biases
Made with ML - Goku Mohandas
Weights & Biases
Angela & Danielle โ Designing ML Models for Millions of Consumer Robots
Weights & Biases
Deep Learning Salon by Weights & Biases
Weights & Biases
More on: ML Maths Basics
View skill โ
Related AI Lessons
โก
โก
โก
โก
Recursion Isnโt Hard. The Call Stack Is Invisible
Dev.to ยท N Satyadev
Understanding Core Probability Concepts in Machine Learning
Medium ยท Machine Learning
Kaggle Getting Started Competition -- Petals to the Metal
Dev.to AI
Understanding the Machine Learning Development Lifecycle (MLDLC)
Medium ยท AI
๐
Tutor Explanation
DeepCamp AI