You may already know it, but in case you don't, in this moment is a trending feature in several startups and companies. Machine Learning is the recognition of patterns using algorithms that learn from statistics models built from big sets of data.
Those algorithms have several applications on industry, some of it are spam filtering, search engines (yes google!), prediction analysis, recommendations systems. Actually they are the current features on social networks like Facebook, video players like Netflix. Yes!, they are all over the software ecosystem, and is not limited to academic fields or University research groups. More even there are several startups that are growing faster because of Machine Learning, some of them are specialized on sell it like a product and were bought by the biggest ones like google or twitter.
However Machine Learning algorithms are not magical potions, as I wrote at the beginning of this post, they learn from big sets of data using math. This is a whole process, it starts mining the data, significant and meaningful data. Choosing the right question based on the mined data. Build an accurate prediction model and teach it to the algorithm. Train the algorithm using training sets from the initial data. Test it with the testing sets. And finally if you don't get what you want, guess what?, may you should start from the beginning.
If you are getting interested in learn it I will share with you some resources:
- Coursera Parctial Machine Learning
- Detailed answer from Quora about How to learn ML
- ML Tutorial and apliances by Nick McCrea
Are you doing machine learning on your Startup share your experience with us.