Multiple machine learning tasks

Details

Supervisor:	Shichang Ding,Please contact sding@gwdg.de or scdingwork@gmail.com
Type:	Student Project (plus potential thesis)
Status:	open

There are multiple machine learning tasks for students, like user profiling, privacy protection, human mobility modeling, recommending system all based on deep learning methods.

What to do: Discuss with me and pick a topic that you are interested in. Most of the topics consist of similar tasks, mainly about implementing features and deep neural network models. Just based on different datasets.

Why important: Machine learning, especially deep learning is one of the hottest topics now. An experience of a real machine learning task will certainly help you to apply for a good position in both famous companies and universities. And if the results are better than state-of-the-art, you will also be the author of an open-published paper.

Requirements: The students' works are mainly about coding. So you need to know how to use Python. As for the knowledge about Machine Learning, if you find one of the following links are not too hard for you, then you can contact me. https://www.kaggle.com/atorin/mnist-digit-recognition-with-random-forests https://keras.io/examples/mnist_cnn/ https://www.kaggle.com/ydalat/titanic-a-step-by-step-intro-to-machine-learning

How to do: I will show you the codes written by me. Then we discuss how to extend or revise them to achieve our targets. Every week I will have a short meeting to help you to solve problems.

The work is predominantly targeted at a 2- or 3-month student project, but can be extended into a Master's thesis as well.