When Machine Learning meets Food Delivery Startups: Ruby Sample Code

http://blog.prediction.io/food-delivery-apps-meet-machine-learning-tutorial-sample-code/#.UkHOZfUFbZs.hackernews

by smhchan • 13 years ago

37 11 13 years ago

blog.prediction.io

berberous• 13 years ago

I wish Seamless would innovate more. Recommended restaurants or meals would be awesome. Seamless has a total lock on the NYC market, but their website sucks, their ratings are highly unreliable (due to restaurants faking ratings), and they rarely add new features.

seiji2• 13 years ago

I wish Seamless would innovate more. is incompatible with Seamless has a total lock on the NYC market, sadly.

I wish you could search intra-menu. "Show me all restaurants with sweet potato fries within walking distance."

Yelp can kinda do that, and Google can kinda do that, but they are just going off reviews and not indexing menus attached to locations specifically.

jeffasinger2• 13 years ago

This is actually a pretty good idea. Seamless does have a program that allows selected affiliates to get data dumps from them, you could probably build it based on that.

pnachbaur• 13 years ago

Is this entirely an open source project, or are there plans to do something like Keen.io?

tstonez2• 13 years ago

Not so familiar with Keen IO. Looks like analytics-as-a-service for collecting and visualizing data. We're focused on helping developers build predictive features like recommendation, discovery, etc.

Yup. We're definitely committed to open source. Source is all up on GitHub https://github.com/predictionio

fancy_pantser• 13 years ago

There is no data provided here. It's just a company blog saying how its product can be used.

gniquil• 13 years ago

So I just looked at the quick start demo code. Did everyone notice the irony in it? We created some fake people and some fake orders and randomly let the people see the fake orders. Then we try to predict based on this data. A bit silly isn't it?

But anyway, the whole thing seems pretty cool.

tstonez2• 13 years ago

Best fake recommendations ever! But yeah we make the assumption the developer and their app (food delivery or otherwise) has some data (users, items, actions) on which to build a model. Otherwise you won't get very far...

Thanks! Some of our tutorials have test datasets (e.g., MovieLens) if you want to try it out - http://docs.prediction.io/current/tutorials/movie-recommenda...

jalayir• 13 years ago

Get a dabbawalla

jeffasinger2• 13 years ago

This is cool. I may try to do this (food ordering predictions) for my startup, with real data if I get some free time.

When Machine Learning meets Food Delivery Startups: Ruby Sample Code

Related Stories