| CS290D - Schedule | home | schedule | project |
| Weeks | Topics | Presenter | Notes | |
| Week 1 (03/30/2015) | Introduction | |||
| Week 1 (04/1/2015) | Neural Networks | Fangqiu Han | ||
| Week 2 (04/6/2015) | Convolutionary Neural Networks | Fangqiu Han |
http://www.nlpr.ia.ac.cn/cip/liukang.files/coling2014.pdf http://web.eecs.umich.edu/~honglak/icml09-ConvolutionalDeepBeliefNetworks.pdf |
|
| Week 2 (04/8/2015) | Word2Vec, GloVe | Yu Su |
Word2vec:
https://drive.google.com/file/d/0B7XkCwpI5KDYRWRnd1RzWXQ2TWc/edit Glove Video: https://www.youtube.com/watch?v=RyTpzZQrHCs&feature=youtu.be Another related tutorial: http://emnlp2014.org/tutorials/8_notes.pdf |
|
| Week 3 (04/13/2015) | More word representations | Yu Su |
https://www.cs.ox.ac.uk/files/6605/aclVectorTutorial.pdf http://www.akbc.ws/2014/slides/weston-nips-akbc.pdf |
|
| Week 3 (04/15/2015) | Question Answering | Huan Sun |
To merge slides from Stanford and Berkeley: http://spark-public.s3.amazonaws.com/nlp/slides/qa.pdf http://www.cs.berkeley.edu/~klein/cs288/sp10/slides/SP10%20cs288%20lecture%2024%20--%20question%20answering%20(2PP).pdf |
|
| Week 4 (04/20/2015) | NLP | Yang Li | Slides from https://www.fer.unizg.hr/_download/repository/TAR-02-NLP.pdf | |
| Week 4 (04/22/2015) | Knowledge Base | Yang Li | Slides from http://resources.mpi-inf.mpg.de/yago-naga/vldb2014-tutorial/vldb2014-slides.pdf | |
| Week 5 (04/27/2015) | Paraphrasing | Izzedine Gur | Slides from http://ir.hit.edu.cn/~zhaosq/paper/PA.pdf | |
| Week 5 (04/29/2015) | More Question Answering | Huan Sun | proposal due | Summarization of recent main methodologies for QA, including open IE, semantic parsing, graph querying, feature-based, and embedding methods. Will prepare slides based on papers from UW, Stanford, JHU, UCSB, Facebook&MSRA ( http://emnlp2014.org/papers/pdf/EMNLP2014071.pdf ; http://emnlp2014.org/papers/pdf/EMNLP2014067.pdf ) |
| Week 6 (05/4/2015) | More NLP | Yang Li | Entity extraction/linking, Relation extraction | |
| Week 6 (05/6/2015) | More Knowledge Base | Yang Li | Slides from http://www.cs.technion.ac.il/~gabr/publications/papers/KDD14-T2-Bordes-Gabrilovich.pdf | |
| Week 7 (05/11/2015) | Deep Learning | Fangqiu Han | Slides made from coursera and Deep learning tutorial in ICML 13' http://www.cs.nyu.edu/~yann/talks/lecun-ranzato-icml2013.pdf | |
| Week 7 (05/13/2015) | Midterm | Midterm will cover materials taught till (including) May 4, 2015. | ||
| Week 8 (05/18/2015) | Recurrent Neural Networks | Semih Yavuz | Slides from second chapter of the tutorial, namely "Deep Learning for NLP without Magic", by Richard Socher: http://nlp.stanford.edu/courses/NAACL2013/NAACL2013-Socher-Manning-DeepLearning.pdf, and maybe introduce a recent paper "A Neural Network for Factoid Question Answering over Paragraphs" (http://cs.umd.edu/~miyyer/pubs/2014_qb_rnn.pdf) as a concrete work. | |
| Week 8 (05/20/2015) | Recursive Neural Networks | Semih Yavuz | For the general overview of RNNs for sentiment analysis, I will use the following three sources: 1) A Note by Prof. Charles Elkan: http://cseweb.ucsd.edu/~elkan/250B/learningmeaning.pdf, 2) My own project report: https://www.academia.edu/6634023/Predicting_Sentiment_Distributions_with_Semi-Supervised_Recursive_Autoencoders, 3) Slides from the same source above. For a concrete work for which RNN approach achieved successful results, I will use the paper "Semi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions" by Richard Socher (http://www.socher.org/uploads/Main/SocherPenningtonHuangNgManning_EMNLP2011.pdf). | |
| Week 9 (05/25/2015) | Holiday | |||
| Week 9 (05/27/2015) | Neural Models with Memory | Jian Shi | ||
| Week 10 (06/1/2015) | Project presentation/demo | |||
| Week 10 (06/3/2015) | Project presentation/demo | |||
| Week 11 (06/7/2015) | No Class | Final report due |