In recent years, machine learning (or data science) has really been taking off, moving beyond academia, and being applied in lots of different fields. Here are a few real-world interesting applications I’ve come across:
- From Amnesty International, a preliminary project to predict which incidents are likely to escalate. Using linguistic information from Amnesty’s logs of historical events, volunteers were able to predict which incidents were likely to escalate, providing a possible way to target more effectively and address situations before they become critical.
- Finding out how and why people donate money in the wake of disasters. People react emotionally to certain words and phrases when disasters happen. Knowing what these are enable aid agencies to report effectively and get money to where it’s needed most.
- Giving a voice back to those with speech impairments – using speech technology to enable those who have problems speaking to communicate more clearly.
- IBM’s Watson (the computer that beat humans at the game Jeopardy) has been tasked with inventing new recipes.
- Using genetic profiles of tumours to help develop new drugs and fight cancer.
- Computer generated singing, can synthesise sounds that are song-like but impossible for humans to produce.