The principle of suggesting interesting users or websites can be also used for journey planning. Every traveler (and runner by the way) has individual preferences. Some want to travel fast, others comfortable, some prefer nice ways with points of interest, some want safe money. machines can learn from users and support to find the right way.
railwaygazette.com - Machine learning to customize journey planning
That sounds nice, but there will be days you need to change your preferences. In fact it is pretty to confuse an algorithm which is based on historical data only.
And what's valid for wearables is also valid for concierge systems. We need prediction on basis of past, present data and future forecasts.
Keine Kommentare:
Kommentar veröffentlichen