{ "id": 174536, "title": "Probabilistic Machine Learning: A Different Approach!", "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/174536", "author_name": "Carlos Souza", "post_date": "2020-08-14T02:28:48.479000", "votes": 14, "comment_count": 5, "views": 0, "content": "

Hi all,
\nI've been silent over the past weeks, learning new tricks involving Probabilistic Machine Learning, Bayesian Networks/Inference, Probabilistic Programming Languages, etc.

\n

These tools are very cool and look like a good approach to model the tabular data, as they are widely used in personalized medicine, and the go-to approach when uncertainty plays a big role in the problem we are tackling. Anyway, I've created a notebook demonstrating the approach. As I'm learning it, would love to get some feedback:

\n

Probabilistic Machine Learning: A Different Approach

\n

In that, I implemented the following Bayesian Network:

\n

\"drawing\"

\n

And here's a demonstration of some of its predictions:

\n

\"drawing\"

\n

More details on the notebook.. Cheers!

", "messages": [ { "id": 969856, "postDate": "2020-08-14T02:28:48.480Z", "content": "

Hi all,
\nI've been silent over the past weeks, learning new tricks involving Probabilistic Machine Learning, Bayesian Networks/Inference, Probabilistic Programming Languages, etc.

\n

These tools are very cool and look like a good approach to model the tabular data, as they are widely used in personalized medicine, and the go-to approach when uncertainty plays a big role in the problem we are tackling. Anyway, I've created a notebook demonstrating the approach. As I'm learning it, would love to get some feedback:

\n

Probabilistic Machine Learning: A Different Approach

\n

In that, I implemented the following Bayesian Network:

\n

\"drawing\"

\n

And here's a demonstration of some of its predictions:

\n

\"drawing\"

\n

More details on the notebook.. Cheers!

", "rawMarkdown": "Hi all,\nI've been silent over the past weeks, learning new tricks involving Probabilistic Machine Learning, Bayesian Networks/Inference, Probabilistic Programming Languages, etc.\n\nThese tools are very cool and look like a good approach to model the tabular data, as they are widely used in personalized medicine, and the go-to approach when **uncertainty** plays a big role in the problem we are tackling. Anyway, I've created a notebook demonstrating the approach. As I'm learning it, would love to get some feedback:\n\n[Probabilistic Machine Learning: A Different Approach](https://www.kaggle.com/carlossouza/probabilistic-machine-learning-a-diff-approach)\n\nIn that, I implemented the following Bayesian Network:\n\n\"drawing\"\n\nAnd here's a demonstration of some of its predictions:\n\n\"drawing\"\n\nMore details on the notebook.. Cheers!", "votes": 14 }, { "id": 969986, "postDate": "2020-08-14T05:25:06.373Z", "content": "

Looks very interesting! I look forward to seeing your notebook

", "rawMarkdown": "Looks very interesting! I look forward to seeing your notebook" }, { "id": 969953, "postDate": "2020-08-14T04:53:57.540Z", "content": "

This seems promising. I have also been inactive; learning new techniques, but I didn't know bayesian networks would be a good choice, I'll be waiting for the results ;)

", "rawMarkdown": "This seems promising. I have also been inactive; learning new techniques, but I didn't know bayesian networks would be a good choice, I'll be waiting for the results ;)", "replies": [ { "id": 970605, "postDate": "2020-08-14T15:47:29.823Z", "content": "

@aadhavvignesh where are you learning it from and what exactly because even I am confused on how to tackle this uncertainty in predictions 😅

", "rawMarkdown": "@aadhavvignesh where are you learning it from and what exactly because even I am confused on how to tackle this uncertainty in predictions 😅", "votes": 1 }, { "id": 970641, "postDate": "2020-08-14T16:19:33.833Z", "content": "

Not a certain source, but I try to find something related to the current techniques by searching on the internet, or eventually end up with a new technique. Browsing for more techniques would help in the competition :)

", "rawMarkdown": "Not a certain source, but I try to find something related to the current techniques by searching on the internet, or eventually end up with a new technique. Browsing for more techniques would help in the competition :)", "votes": 1 } ] }, { "id": 972586, "postDate": "2020-08-16T16:56:15.050Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 969986, "author_name": "Matt", "author_url": "", "post_date": "2020-08-14T05:25:06.373000", "content": "

Looks very interesting! I look forward to seeing your notebook

", "votes": 0, "replies": [] }, { "id": 969953, "author_name": "Aadhav Vignesh", "author_url": "", "post_date": "2020-08-14T04:53:57.540000", "content": "

This seems promising. I have also been inactive; learning new techniques, but I didn't know bayesian networks would be a good choice, I'll be waiting for the results ;)

", "votes": 0, "replies": [ { "id": 970605, "author_name": "Digvijay Yadav", "author_url": "", "post_date": "2020-08-14T15:47:29.823000", "content": "

@aadhavvignesh where are you learning it from and what exactly because even I am confused on how to tackle this uncertainty in predictions 😅

", "votes": 1, "replies": [] }, { "id": 970641, "author_name": "Aadhav Vignesh", "author_url": "", "post_date": "2020-08-14T16:19:33.833000", "content": "

Not a certain source, but I try to find something related to the current techniques by searching on the internet, or eventually end up with a new technique. Browsing for more techniques would help in the competition :)

", "votes": 1, "replies": [] } ] }, { "id": 972586, "author_name": "", "author_url": "", "post_date": "2020-08-16T16:56:15.050000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "969856": "Hi all,\nI've been silent over the past weeks, learning new tricks involving Probabilistic Machine Learning, Bayesian Networks/Inference, Probabilistic Programming Languages, etc.\n\nThese tools are very cool and look like a good approach to model the tabular data, as they are widely used in personalized medicine, and the go-to approach when **uncertainty** plays a big role in the problem we are tackling. Anyway, I've created a notebook demonstrating the approach. As I'm learning it, would love to get some feedback:\n\n[Probabilistic Machine Learning: A Different Approach](https://www.kaggle.com/carlossouza/probabilistic-machine-learning-a-diff-approach)\n\nIn that, I implemented the following Bayesian Network:\n\n\"drawing\"\n\nAnd here's a demonstration of some of its predictions:\n\n\"drawing\"\n\nMore details on the notebook.. Cheers!", "969986": "Looks very interesting! I look forward to seeing your notebook", "969953": "This seems promising. I have also been inactive; learning new techniques, but I didn't know bayesian networks would be a good choice, I'll be waiting for the results ;)", "972586": "" } }