Scary Mathematics

Halloween is tomorrow (for me), so I think that posting about mathematics related to Halloween will be very interesting

1. Carving pumpkin. This is one of my favorite Halloween traditions. Romania (my place of birth) did not celebrate, but in autumn pumpkins were carved for my grandparents and my sister. This is some good ideas for mathematicians:

A terrible mathematics. In the history of mathematics, there are strange names for specific attributes and specific numbers. Among these names are those that are perfect for Halloween; they will remind you of witches, beasts, demons, skeletons etc unique to October 31. This is the best (and most interesting) thing I think.

clothes. This chapter is for those who love Halloween. I think dressing up for Halloween is really creative. If you are like a more mathematic boyfriend (like a person in the image - the image is not mine, Google will help me), it's a bit strange and fun to dress up this Halloween There is a way:

Thank you for your support and cooperation. As any feedback will be evaluated, please click on the like button. In addition, you can use # HalloweenwithLThMath to send me pictures of what you are doing for Halloween. Even if it is not related to mathematics.

My Facebook page, my Tumblr, the Google+ page I just started, and my new Twitter (I am familiar with the two recent things if you saw those stupid mistakes). Thank you for reading and enjoying. Have a nice weekend

Return to the story before the deep learning era - when the neural network is terrible than strange machine learning and artificial intelligence tools, strange mathematical curiosity - there are lots of classically successful data mining that has surprisingly succeeded in natural language processing Algorithm application (NLP) domain name. To deal with problems such as spam filtering and part-of-speech tagging, you can use models which are fairly simple and easy to understand. However, not all problems can be solved in this way. With a simple model, it is not possible to properly capture subtle parts of the context such as context, idioms, satire (although at that point human beings will fail). Algorithms based on the general summary (such as word porch) have proved to be insufficient to capture a series of text data, but n-gram is difficult to simulate a general background , I am greatly influenced by the dimension curse

Natural language processing in artificial intelligence is almost human's hand. Worse, it gets smarter!

For the first few years it was a devil that was used to solve mathematics and physics. At that time, even though I was not graduating from college, I thought it was very difficult to spend time with my friends. Because I was always afraid to return to these subjects. But thanks to omnipotent things, things are going well and there is no backup. The situation has changed when I entered the last semester of the course. Due to the emergence of core computer subjects, research tension is not high. Because these things are natural to me. Anyway, this is what I started going to a cafe with my friends.