comment-board

  1. shouldsee,feng.geng.14"AT"ucl.ac.uk : wordpress revived after a ton of tinkering.

  2. shouldsee,feng.geng.14"AT"ucl.ac.uk : What works for biologists, may not work for computer scientists.

  3. shouldsee,feng.geng.14"AT"ucl.ac.uk : In reply to Bullet51. BTW, I don’t seems to get email on this blog. Maybe github is a better place to talk if you can reach it. (https://github.com/MathAndAlgo/DiscussionBoard/issues).

  1. shouldsee,feng.geng.14"AT"ucl.ac.uk : In reply to Bullet51. That video looks really intriguing. You know what, the protein structure people was doing exactly the same thing! They inferred a empirical force-field from a collection of proteins and applied it to get sensible protein structures. (See https://salilab.org/modeller/)

  2. shouldsee,feng.geng.14"AT"ucl.ac.uk : In reply to Bullet51. Haven’t spoken for long! Are you still studying GOL recently?

  3. Bullet51,w_w3333"AT"126.com : Particle Life. Looks like Biochem. https://www.youtube.com/watch?v=Z_zmZ23grXE https://github.com/HackerPoet/Particle-Life

  4. shouldsee: Odour NMF https://flavourjournal.biomedcentral.com/articles/10.1186/2044-7248-3-S1-P3

  5. A full introduction to probabilistic PCA by Lior Patcher. https://liorpachter.wordpress.com/2014/05/26/what-is-principal-component-analysis/

  6. In reply to shouldsee. Inverse ising problems and probability flow https://arxiv.org/abs/0906.4779

https://www.researchgate.net/publication/322713743_Convenient_Interface_to_Inverse_Ising_ConIII_A_Python_package_for_solving_maximum_entropy_models

  1. Animals: The loyal geese: https://www.peta.org/blog/geese-kiss-goodbye/

  2. In reply to shouldsee. Image denoising by anisotropic diffusion: http://people.maths.ox.ac.uk/capdeboscq/HerbertAmann-Oxford-150609.pdf

  3. Graphics: R: http://zoonek2.free.fr/UNIX/48_R/04.html 中文生信社区: http://www.bio-info-trainee.com http://www.biotrainee.com:8080/

  4. 典故集:赵襄王学御/心不在马 ——《韩非子·喻老》 寧戚 齊桓公 淮南子

  5. 今日思考:如何回应一个不想和你谈话的人发出的谈话意向? 互补问题:如何在交流的时候避免对他人的轻视?

  6. 友情链接:Rosetta

  7. Useful Applied Stat Slides from ETH https://stat.ethz.ch/education/semesters/ss2012/ams/slides/v7.2.pdf “hmmscan VS hmmsearch” : https://cryptogenomicon.org/2011/05/27/hmmscan-vs-hmmsearch-speed-the-numerology/ Difference between logarithm of an expectation value and expectation value of a logarithm: https://math.stackexchange.com/questions/21063/difference-between-logarithm-of-an-expectation-value-and-expectation-value-of-a

Reinforcement learning:http://karpathy.github.io/2016/05/31/rl/ Williams1992, REINFORCE algorithm: http://www-anw.cs.umass.edu/~barto/courses/cs687/williams92simple.pdf Andrej Karapathy writes excellent blogs for neural networks. http://karpathy.github.io/ RNN and HDM: http://slideplayer.com/slide/8844792/ Cosma Shalizi 元胞自动机大神的非线性filtering http://bactra.org/weblog/375.html Some postdoc from DAMTP https://rajeshrinet.github.io/blog/ Denoising: a diffusion-based approach http://people.maths.ox.ac.uk/capdeboscq/HerbertAmann-Oxford-150609.pdf Inverse ising problems and probability flow https://arxiv.org/abs/0906.4779 https://www.researchgate.net/publication/322713743_Convenient_Interface_to_Inverse_Ising_ConIII_A_Python_package_for_solving_maximum_entropy_models

Regex 配合 StringIO 简直Hacking神器 未来和过去的分界线,是当下。 Web icloud 用户体验差出了新高度。

Deconvolution: https://terpconnect.umd.edu/~toh/spectrum/Deconvolution.html Image denoising by anisotropic diffusion: http://people.maths.ox.ac.uk/capdeboscq/HerbertAmann-Oxford-150609.pdf

  1. Today’s Point: Periodicity often coincides with complexity, but NOT always! Breeders can successfully manage far more complex genetic interactions via phenotypic selection among fixed lines than even the most recent molecular approaches can explain. A breeder with a large, multi-national breeding and chemical company was recently overheard to say that they would have made much greater breeding progress if they hadn’t been so ‘distracted‘ by molecular biology. There can be too much information, and generally not enough relevant knowledge. “(Falk, Duane E. 2008)

  2. 一个问题: Back propagation如果把所有output给出的gradient进行平均会给出什么样的动力学? 关于动力系统的那些奇技淫巧。「降维攻击」 https://sinews.siam.org/Details-Page/finding-emergent-low-dimensional-macroscopic-behavior-in-large-systems-of-many-coupled-dynamical-units#

  3. 晚上十点以后不编程,不算积分,不看电影。

  4. Markovian Ising

  1. 2019-09-26 What does it mean to be friendly or not?

  2. Gooses Geese kiss-goodbye

  3. python packaging: setup.py

I have been struggling with python2 setup.py install not installing the package correctly even an “package.egg” is created under “/usr/local/lib/python2.7/dist-packages”. After 2hr of searching I found out a simple solution:

If you have a can’t iXXXrt XXX error. Try use disutils.core.setup() instead of setuptools.setup() in your “setup.py”

  1. Bioinf_diary:
  1. url_collection

曹刿与范文正公

乡人:“肉食者谋之,又何间焉?”

刿:“肉食者鄙,未能远谋。”

范文正公:“余常求古仁人之志,或异二者之为。居庙堂之高则忧其民,处江湖之远则忧其君。是进亦忧,退亦忧。然则何时而乐耶?其必曰:“先天下之忧而忧,后天下之乐而乐”乎。噫!微斯人,吾谁与归。”

现在问题来了,当刿碰上范,会擦出怎样的火花呢?

Comments