Information consumption is readily done on a per item basis,
Those of us that keep our documents and email sorted and ordered – that’s a plus.
Most of us, however, do not know how find what they want, even if they made some sort of order and performed a search.
Moreover, many more of us could benefit a “quick memory recharge” by consuming information on a per topic basis and have it multi-dimensionally ordered and controlled to fit the time line and encompassing topics
My master’s thesis titled:
Building a diverse focused library of HSP90α inhibitors using the WOBL fuzzy matcher
Will be presented tomorrow the 31th of December 2008 at 15:00 in the Hebrew university, school of pharmacy, hall A.
You are more than welcome to come…
I haven’t been posting anything for almost a year now,
I have been intensively preoccupied with my research, which is now coming to its completion, at least from the academic point of view…
I’ll continue writing soon, and now, hopefully, in a more consistent manner.
Cheminformatics and mining quotation contains quotes of famous scientists regarding data mining, drug discovery and chemoinformatics. Each quotation comes from an article, for example, John H. VanDrie’s (1995) article on 3D Database Searching in Drug Discovery: “You’re always explaining to me after the fact why a molecule I’ve already discovered is active, but you’re never able to tell me ahead of time which molecules I should be investigating.”
The rules of chemical bonding allow simple elements such as carbon, hydrogen, oxygen, nitrogen and fluorine to form a vast multitude of different molecules. This so-called chemical universe or chemical space probably contains 10^18-10^200 molecules with up to 30 atoms. Up to now however nobody had attempted to see which structures populate chemical space. We have taken such a first look by constructing a database of all molecules up to 11 atoms under constraints that define chemical stability and synthetic feasibility. The database contains 26.4 million compounds, the vast majority of which have never been synthesized. Searching through this database might help to identify new synthetic TARGETs with interesting properties, and the discovery of new drugs. An earlier version of this database has been reported. The present version includes molecules with small rings (3 and 4-membered rings) and triple bonds.
academic license agreement here
Bayesian theorem enables calculating probability dependence (inference) and enables bias correction:
Moreover, one can calculate the change in probability A when B occurs and vice verse. A silly yet understandable example: the probability that one will get wet when it rains is different than of one getting wet due to rain. Additional feature of the above theorem is the ability to “correct” the calculated results according to the observed data.
Occam’s razor, states that “entia non sunt multiplicanda praeter necessitatem” which translates to: “entities should not be multiplied beyond necessity.” saying: simplicity is the evidence of truth.
Going further to Information theory, the “minimum description length principle” aka MDL, formalizes Occam’s razor in stating that the best hypothesis for a given set of data is the one that leads to the largest compression of the data. It also provides a natural safeguard against overfitting.
A proper implementation will result, that searching for an efficient code reduces to searching for a good probability distribution, and vice versa.
For further reading Bayesian publications in statistics might supply additional insights
An article by freeman dyson on the need of heretics in science
Capability Maturity Model Integration at wikipedia
IBM Developerworks’ Jeff Vogel explains six methods for writing clean and understandable code.
LiFusion 
You must be logged in to post a comment.