01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
2008 : 01 02 03 04 05 06 07 08 09 10 11 12
2007 : 01 02 03 04 05 06 07 08 09 10 11 12
2006 : 01 02 03 04 05 06 07 08 09 10 11 12
2005 : 01 02 03 04 05 06 07 08 09 10 11 12
<http://www.springerlink.com/content/v6mn41027420jp56/>
http://www.cs.jhu.edu/~junwu/publications.html
文脈中のhead word素性、近接Nグラム素性(通常のNグラムに相当)、単語のクラス素性、
SVMでもちいられている多くのカーネルと、
MEなどでもちいられている組み合わせ素性を比較した場合、
カーネルは可読性が低い。
MEの組み合わせ素性は、それの重みをみることで簡単にどの組み合わせが有効かをみることができる。
一方、カーネル法の場合、カーネル関数のどの項をけずった場合に性能が劣化するかを見る、
などの間接的な方法しかとれない。
これを解決するため、すべての組み合わせを列挙することに相当するpower set kernelと
その高速な計算法を提案した。
<http://valgrind.org/info/tools.html>
Official Home Page for valgrind, a suite of tools for debugging and profiling. Automatically detect memory management and threading bugs, and perform detailed profiling. The current stable version is valgrind-3.3.0.
キャッシュ使用状況のエミュレーションと回数つきコールグラフの生成 callgrind
とその可視化 kcachegrind
<http://hagi.is.s.u-tokyo.ac.jp/ade/pre/1.html>
課題1-A: シェルを実装せよ。
課題2-A: データを圧縮・解凍するプログラムを実装せよ。
課題3-A: スパムフィルタを実装せよ。
この課題が授業の目標:伝えられた仕様を大体何でも実装できるようになるこ
とという風に解釈した。
詳しくは別紙
面白いのは圧縮解凍、スパムフィルタに関しては、
圧縮率、正解率という優劣をかなりはっきり決める一次元の指標があること。
(シェルは基本は誰が作っても同じ、詳細な機能の設計に差異が出るが優劣は個別にみる必要がある)
<http://jaortega.wordpress.com/2007/01/31/a-scheme-bookshelf/>
yet another introduction to (functional) programming
<http://nlp.cs.swarthmore.edu/sigphon/index.php>
ACL 2008 でワークショップ
http://phonology.cogsci.udel.edu/sigmorphon2008/
Submission Deadline March 17, 2008, 23:59 EST
産総研実習を出したい。
Pastry - A scalable, decentralized, self-organizing and fault-tolerant substrate for peer-to-peer applications
2008 : 01 02 03 04 05 06 07 08 09 10 11 12
2007 : 01 02 03 04 05 06 07 08 09 10 11 12
2006 : 01 02 03 04 05 06 07 08 09 10 11 12
2005 : 01 02 03 04 05 06 07 08 09 10 11 12
最終更新時間: 2009-02-01 00:57
Powered by chalow