The Interpretable Machine Learning Podcast
Technology Podcasts
This podcast is dedicated to Interpretable Machine Learning - a branch within the wider field of Machine Learning which aims to make the inner-workings of models easier for humans to understand. This is a popular and fast moving area of research which also has many important applications in industry. However, there isn't very much podcast content available on Interpretable Machine Learning. By interviewing researchers and practitioners, I hope to address this shortcoming with an interesting and informative show!
Location:
United States
Genres:
Technology Podcasts
Description:
This podcast is dedicated to Interpretable Machine Learning - a branch within the wider field of Machine Learning which aims to make the inner-workings of models easier for humans to understand. This is a popular and fast moving area of research which also has many important applications in industry. However, there isn't very much podcast content available on Interpretable Machine Learning. By interviewing researchers and practitioners, I hope to address this shortcoming with an interesting and informative show!
Language:
English
Website:
https://anchor.fm/neil-gibbons2
Episodes
#5 Sara Hooker: Interpreting Deep Learning Models
10/19/2021
In this episode, I speak to Sara Hooker of Google Brain about her work in interpreting deep learning models.
Duración:00:49:52
#4 Alex Luketa: Interpretable Machine Learning at Xerini
9/10/2021
1:18 - Was there a project where it was especially important to build a client’s trust in a model?
3:59 - Has there been a project where you opted for a less complicated model, sacrificing accuracy for interpretability?
8:00 - I notice you’ve used decision trees in a few projects, is interpretability an important factor in your choice of this type of model?
9:58 - Is there an example of a project where interpretability wasn’t actually that important to you? Where you weren’t willing to...
Duración:00:17:35
#3 Shirin Elsinghorst: using IML as a Data Scientist; model-agnostic interpretability methods; and the link between IML and ethics
9/2/2021
1:01 – What is the source of your interest in IML?
2:20 - Has IML become more mainstream in the Data Science profession?
2:54 – Do you use IML in your work as a Data Scientist?
4:17 – What domains do you work in?
5:00 – Do you work with structured or unstructured data, and do you generally use models with interpretability “built-in” to them or more black box models?
6:36 – Are you ever tempted to use complex, exciting and uninterpretable models instead of simpler options?
8:10 – What...
Duración:00:29:08
#2 - Rich Caruana: IML in healthcare, EBMs and the future of IML
7/29/2021
In this episode, I speak to Dr Rich Caruana of Microsoft Research about his work in Interpretable Machine Learning.
2:30 - How did you first become interested in IML?
7:20 - What surprises in the pneumonia dataset did you discover with interpretable methods?
12:12 - Were there any surprises which were not mistakes, but rather helped you to learn something new about the dataset you were working with?
The graph Rich describes in answering this question
16:53 - Can you explain Explainable...
Duración:00:54:56
#1 - What is Interpretable Machine Learning?
7/13/2021
In this short first episode, I introduce the motivation and idea behind the podcast, before giving a high-level introduction to Interpretable Machine Learning.
Duración:00:13:31