mcrowe

Thanks for your thoughts ehllo. I don't happen to agree that this would be a better design. The tool is meant to be useful, not impressive.

Hi hackers! I built NameQL to help me find names for my new products, and I hope you'll find it helpful too.

It considers 100k+ names in milliseconds and shows you the best for which [name].com is available. Sure, tools like this exist, but I don't know of any that come close in terms of speed.

I used an n-gram predictive model, trained on an English word corpus to generate the suggestions. For ridiculous performance, I used a Bloom filter backed by Redis, together with some sneaky and caching and prefetching.

My next plans are to enhance this by searching the app stores for similarly named apps, and also checking for relevant trademarks, since I find myself doing that a lot manually. And I would like to add more advance name generation controls, like choosing prefix words matching parts of speech and using n-gram word models to generate phrase suggestions.

Thanks for checking it out!

Beautiful work! Reminds me of a very polished Jupyter.

Any plans to support compile-to-js languages? Coffeescript, etc.

The source code is now up on github. https://github.com/mcrowe/neurovis

Well said. That's exactly the point: to get "play-with-it level intuition". Even though its not rigorous, I still find it helps.

I am planning on building a simulation of the back-prop algorithm though!

Glad you enjoyed it! It uses a library called vis.js.

Thanks!

Yes. I will be posting the source to github very soon.

Neural networks are a beautiful concept. They are also beautiful to watch. I created an interactive neural network visualizer and tutorial. It’s a fun way to develop some intuition about neural nets. It was created using vis.js, and a lot of love. Hope you enjoy!

Yes. "Ask HN" and "Show HN" both do almost twice as well as standard stories.

Thanks for clarifying, dang. My apologies.

Haha. Actually, Hacker News admin sent me an email asking me to re-submit it with a slightly different URL.

You're right. I was focused on telling a useful story with the data. I should be more careful not to be misleading.

Your related point is a good catch. That was a brain slip on my part. Thanks for pointing it out!

I looked at the data from 1.3 million Hacker News stories and found that when a story gets submitted makes a big difference (up to 172% better chance of getting on the front page). This article shows the analysis and results.

I used the official Hacker News API to get the stories using Python, and used R and ggplot2 to do the exploratory data analysis and plots.

Absolutely! The "success metric" will be different for different analyses, though, which would make that tough. The obvious choice is to look at the mean number of votes for stories submitted at different times. I looked at a binary condition instead: the fraction of stories which get more than N votes. This biases things less towards "super-star" stories which get tons of votes, and focuses on whether a story gets discovered at all.

There is a bunch of analysis I did behind the scenes to motivate this choose. I'll be posting the source code for my analysis, so it will be easy for anyone to critique and compare.

I looked at the data from 1.3 million Hacker News stories and found that when a story gets submitted makes a big difference (up to 172%). This article shows the analysis and results.

I used the official Hacker News API to get the stories using Python, and used R and ggplot2 to do the exploratory data analysis and plots.

Neural networks are a beautiful concept. They are also beautiful to watch. I created an interactive neural network visualizer and tutorial. It’s a fun way to develop some intuition about neural nets. It was created using vis.js, and a lot of love. Hope you enjoy!

I wrote this article to help others wade through the murky waters of learning Data Science. I hope you find it useful!