4CSLL5 - Advanced Computational Linguistics: Machine Learning Techniques in Machine Translation, Speech Recognition and Topic Modelling
Lecturer: Martin Emms
Office: Lower Ground Floor O'Reilly LG 18
Phone: ext. 1542
- Course is about how to have machines learn from data, with emphasis on
techniques which can be applied in Machine Translation, Speech Recognition and Topic Modeling, though they are applicable much more widely than that. The big emphasis is on unsupervised machine learning and its this that is this course's most distinctive trait compared to other modules which involve machine learning.
overview of topics, along with pointers to literature (now includes info on Topic Models)
cheat sheet (9 pages) summarising the bare essentials of what has been covered
- Despite the module name, the material will be just as accessible to ICS students as to CSL students: the number of non-CSL students taking the course last year exceeded the number of CSL students
The following brief overview presentations give a little more information
- Please feel free to get in touch by email or in person if you would like further information about the nature of the course
Finally for the curious:
- for the curious these
notes contain a proof that EM increases probability. This is
absolutely not required or even helpful for the exam , but its
cute if you like that kind of thing. I have yet to figure out how to reasonably incorporate this into the course, but one day ....