Viewing profile — ai_maker

There are API aggregation sites like Mashape that are already working toward this goal. My startup thing is about using learned models for text processing [1]. Quick tests are also…

If we are to cite articles here, let me get one from Quim Monzó (a reputed catalan journalist) [1]. I'm pasting a Google translation directly here: "" Noone can stop it The latest …

Very true, but in the real world where what's urgent goes before what's important, you are often forced to reschedule stuff, and the pen&paper approach clearly is in disadvantage h…

Worse is better [1]. [1] https://www.jwz.org/doc/worse-is-better.html

Ask HN: What value can makers add to ham? I feel I belong to this "new makers" wave and I'm thinking about using Artificial Intelligence to model noise (and cancel it) so as to use…

I would say that hill climbing with random restarts works better because you must first define a cost function in a precise (and concise) symbolic mathematical way. Very much simil…

This is an implementation in Python for the second option, but instead of word frequencies it uses word presence (a binary vector in the end): https://github.com/atrilla/vsmpy

I love Lua and I'm doing AI with it (with a veneer of electronics). What do you have in mind?

Good post! A bunch of mistakes is a bunch of opportunities to improve. It kind of complements Domingos' paper: http://homes.cs.washington.edu/~pedrod/papers/cacm12.pdf

Key to AI is clear. But, which is the challenge? Are we talking about predicting actually? Or is it rather about building a huge case-based reasoning system that can "replicate" kn…

Do you have the gold standard labels of your dataset? Can you ensure that the amount of pos/neg labels is symmetrical? You can heuristically tune the weights of your lexicon to fit…