“If the NYSE, Nas­daq, Amex and OTC are trad­ing 2 Bil­lion shares a day or more , like today, thats $ 500 Mil­lion Dol­lars PER DAY. If there are 260 trad­ing days a year. Thats about 130 Bil­lion dol­lars a year. If vol­umes drop because of the tax. It is still 10s of Bil­lions of dol­lars per year.

Thats real money for the US Trea­sury. Thats also an annual pay­ment towards the next time Wall Street screws up and we have a black swan event that no one planned on.”

“Learn­ing in prob­a­bilis­tic mod­els is often ham­pered by the gen­eral intractabil­ity of the nor­mal­iza­tion fac­tor and its deriv­a­tives. Here we pro­pose a new learn­ing tech­nique that obvi­ates the need to com­pute an intractable nor­mal­iza­tion fac­tor or sam­ple from the equi­lib­rium dis­tri­b­u­tion of the model. This is achieved by estab­lish­ing dynam­ics that would trans­form the observed data dis­tri­b­u­tion into the model dis­tri­b­u­tion, and then set­ting as the objec­tive the min­i­miza­tion of the ini­tial flow of prob­a­bil­ity away from the data distribution.…”

“In the rhi­zomatic model of learn­ing, cur­ricu­lum is not dri­ven by pre­de­fined inputs from experts; it is con­structed and nego­ti­ated in real time by the con­tri­bu­tions of those engaged in the learn­ing process. This com­mu­nity acts as the cur­ricu­lum, spon­ta­neously shap­ing, con­struct­ing, and recon­struct­ing itself and the sub­ject of its learn­ing in the same way that the rhi­zome responds to chang­ing envi­ron­men­tal conditions…”

“Such dis­so­lu­tion of math­e­mat­ics may be occur­ring already as evi­denced by the pro­lif­er­a­tion and growth of fields, par­tic­u­larly since the mid­dle of the last cen­tury, that have con­sid­er­able math­e­mat­i­cal con­tent but whose fac– ulties and pro­fes­sional soci­eties have lit­tle over­lap­ping mem­ber­ship. When dif­fer­ent fields spon­sor dif­fer­ent branches of math­e­mat­ics, then the sub­fields may adopt the cul­tures of their spon­sors, so the branches of math­e­mat­ics even­tu­ally may come to dis­agree over accept­able nota­tion, rigor, idiom, and ter­mi­nol­ogy. Bar­ri­ers among the branches of math­e­mat­ics entail a large oppor­tu­nity cost, so to speak, because pos­si­ble col­lab­o­ra­tions and syn­er­gies may not be realized.”

“In this paper we intro­duce a method for deter­min­ing local inter­ac­tion rules in ani­mal swarms. The method is based on the assump­tion that the behav­ior of indi­vid­u­als in a swarm can be treated as a set of mech­a­nis­tic rules.
The prin­ci­pal idea behind the tech­nique is to vary para­me­ters that define a set of hypo­thet­i­cal inter­ac­tions to min­i­mize the devi­a­tion between the forces esti­mated from observed ani­mal tra­jec­to­ries and the forces result­ing from the assumed rule set. We demon­strate the method by recon­struct­ing the inter­ac­tion rules from the tra­jec­to­ries pro­duced by a com­puter simulation.”

“It has long been known that sci­en­tific out­put pro­ceeds on an expo­nen­tial increase, or more prop­erly, a logis­tic growth curve. The inter­play between effort and dis­cov­ery is clear, and the nature of the func­tional form has been thought to be due to many changes in the sci­en­tific process over time. Here I show a quan­ti­ta­tive method for exam­in­ing the ease of sci­en­tific progress, another nec­es­sary com­po­nent in under­stand­ing sci­en­tific dis­cov­ery. Using exam­ples from three dif­fer­ent sci­en­tific dis­ci­plines — mam­malian species, chem­i­cal ele­ments, and minor plan­ets — I find the ease of dis­cov­ery to con­form to an expo­nen­tial decay. In addi­tion, I show how the pace of sci­en­tific dis­cov­ery can be best under­stood as the out­come of both sci­en­tific out­put and ease of discovery.”

“In the present work we intro­duce a novel multi-​​agent model with the aim to repro­duce the dynam­ics of a dou­ble auc­tion mar­ket at micro­scopic time scale through a faith­ful sim­u­la­tion of the match­ing mechan­ics in the limit order book. The model fol­lows a “zero intel­li­gence” approach where the actions of the traders are related to a sto­chas­tic vari­able, the mar­ket sen­ti­ment, which we define as a mix­ture of pub­lic and pri­vate infor­ma­tion. The model, despite the par­si­mo­nious approach, is able to repro­duce sev­eral empir­i­cal fea­tures of the high-​​frequency dynam­ics of the mar­ket microstruc­ture not only related to the price move­ments but also to the depo­si­tion of the orders in the book.”

“We describe a search for prim­i­tive tri­no­mi­als of high degree and its inter­ac­tion with the Great Inter­net Mersenne prime search (GIMPS). The search is com­plete for tri­no­mi­als whose degree is the expo­nent of a Mersenne prime, for all 47 cur­rently known Mersenne primes.”

“Inves­ti­gat­ing the rela­tion between the struc­ture and behav­ior of com­plex bio­log­i­cal net­works often involves pos­ing the fol­low­ing two ques­tions: Is a hypoth­e­sized struc­ture of a reg­u­la­tory net­work con­sis­tent with the observed behav­ior? And can a pro­posed struc­ture gen­er­ate a desired behav­ior? Answer­ing these ques­tions pre­sup­poses that we are able to test the com­pat­i­bil­ity of net­work struc­ture and behav­ior. We cast these ques­tions into a para­me­ter search prob­lem for qual­i­ta­tive mod­els of reg­u­la­tory net­works…. We develop a method based on sym­bolic model check­ing that avoids enu­mer­at­ing all pos­si­ble param­e­triza­tions, and show that this method per­forms well on real bio­log­i­cal prob­lems, using the IRMA syn­thetic net­work and bench­mark exper­i­men­tal data sets. We test the con­sis­tency between the IRMA net­work struc­ture and the time-​​series data, and search for para­me­ter mod­i­fi­ca­tions that would improve the robust­ness of the exter­nal con­trol of the sys­tem behavior.”

“All rolled into one big ball, the biggest thing to take away from this post is to find the job that will make you happy. These are all just things that I have that make me happy, so maybe they’ll help you find that great place to work. Because of all these rea­sons and prob­a­bly some oth­ers I’ll think of after pub­lish­ing this post, thought­bot has my heart. Bar­ring any­thing very unex­pected, and until I’ve got­ten sick of design, you’ll find me here at my desk inside thought­bot HQ. I can only hope you have the same lux­ury or soon find a place that makes you just as happy.”

“Soft­ware projects are dis­played by Gource as an ani­mated tree with the root direc­tory of the project at its cen­tre. Direc­to­ries appear as branches with files as leaves. Devel­op­ers can be seen work­ing on the tree at the times they con­tributed to the project.”

“As a web devel­oper you really have no excuse for not know­ing every­thing there is to know about URLs, there is just not that much to them. But, I have found that even expe­ri­enced devel­op­ers often have some glar­ing holes in their knowl­edge of URLs. So, I thought I would do a quick tour of every­thing that every devel­oper should know about URLs. Strap your­self in – this won’t take long .”

“Our find­ings sug­gest that in ravens, bystanders may con­sole vic­tims with whom they share a valu­able rela­tion­ship, thus alle­vi­at­ing the vic­tims’ post-​​conflict dis­tress. Con­versely vic­tims may affil­i­ate with bystanders after a con­flict in order to reduce the like­li­hood of renewed aggres­sion. These results stress the impor­tance of rela­tion­ship qual­ity in deter­min­ing the occur­rence and func­tion of post-​​conflict inter­ac­tions, and show that ravens may be sen­si­tive to the emo­tions of others.”

“The occur­rence of the nonzero left­most digit, i.e., 1, 2, …, 9, of num­bers from many real world sources is not uni­formly dis­trib­uted as one might naively expect, but instead, the nature favors smaller ones accord­ing to a log­a­rith­mic dis­tri­b­u­tion, named Benford’s law. We inves­ti­gate three kinds of widely used phys­i­cal sta­tis­tics, i.e., the Boltzmann-​​Gibbs (BG) dis­tri­b­u­tion, the Fermi-​​Dirac (FD) dis­tri­b­u­tion, and the Bose-​​Einstein (BE) dis­tri­b­u­tion, and find that the BG and FD dis­tri­b­u­tions both fluc­tu­ate slightly in a peri­odic man­ner around the Ben­ford dis­tri­b­u­tion with respect to the tem­per­a­ture of the sys­tem, while the BE dis­tri­b­u­tion con­forms to it exactly what­ever the tem­per­a­ture is. Thus the Benford’s law seems to present a gen­eral pat­tern for phys­i­cal sta­tis­tics and might be even more fun­da­men­tal and pro­found in nature. Fur­ther­more, var­i­ous ele­gant prop­er­ties of Benford’s law, espe­cially the man­tissa dis­tri­b­u­tion of data sets, are discussed.”

“As the study of infor­ma­tion pro­cess­ing in liv­ing cells moves from indi­vid­ual path­ways to com­plex reg­u­la­tory net­works, math­e­mat­i­cal mod­els and sim­u­la­tion become indis­pens­able tools for ana­lyz­ing the com­plex behav­ior of such net­works and can pro­vide deep insights into the func­tion­ing of cells. The dynam­ics of gene expres­sion, for exam­ple, can be mod­eled with Boolean net­works (BNs). These are math­e­mat­i­cal mod­els of low com­plex­ity, but have the advan­tage of being able to cap­ture essen­tial prop­er­ties of gene-​​regulatory net­works. How­ever, cur­rent imple­men­ta­tions of BNs only focus on dif­fer­ent sub-​​aspects of this model and do not allow for a seam­less inte­gra­tion into exist­ing pre­pro­cess­ing pipelines. RESULTS: Bool­Net effi­ciently inte­grates meth­ods for syn­chro­nous, asyn­chro­nous and prob­a­bilis­tic BNs. This includes recon­struct­ing net­works from time series, gen­er­at­ing ran­dom net­works, robust­ness analy­sis via per­tur­ba­tion, Markov chain sim­u­la­tions [&c]”

“What is it that math­e­mati­cians accomplish?”

“Com­pu­ta­tional Com­plex­ity stud­ies the power and lim­i­ta­tions of effi­cient com­pu­ta­tion. So is effi­cient com­pu­ta­tion purely an abstract math­e­mat­i­cal object or is it try­ing to model a real world phe­nom­e­non? I would argue the lat­ter. Effi­cient com­pu­ta­tion occurs not just in com­put­ers but in bio­log­i­cal sys­tems, phys­i­cal sys­tems, chem­i­cal sys­tems, eco­nomic sys­tems and much more. Physics focuses on the “what”, com­pu­ta­tional com­plex­ity on the “how”.”

“The under­ly­ing real­ity of a suc­ces­sion of inter­breed­ing pop­u­la­tions is a vastly com­pli­cated net­work N. Since Dar­win, species trees have been used as a sim­pli­fied descrip­tion of the rela­tion­ships which sum­ma­rize the overly com­pli­cated net­work N. Recent evi­dence of hybridiza­tion and lat­eral gene trans­fer, how­ever, sug­gest that there are sit­u­a­tions where trees are inad­e­quate. Con­se­quently it is impor­tant to deter­mine prop­er­ties that char­ac­ter­ize net­works closely related to N and pos­si­bly more com­pli­cated than trees but lack­ing the full com­plex­ity of N.”

“Pay atten­tion to the path the steep­est descent takes until find­ing the opti­mum at x = 1 and y = 1. It is said that an evo­lu­tion strat­egy would fol­low a sim­i­lar path if its pop­u­la­tion could be infi­nite. Since, so far, there are no real com­put­ers with infi­nite mem­ory, such assump­tion can­not be ver­i­fied in the real world.”

“This is a guide to the R pro­gram­ming language.”

“Micro­bial ecosys­tems have been widely used in indus­trial pro­duc­tion, but the inter-​​relationships of organ­isms within them haven’t been com­pletely clar­i­fied due to com­plex com­po­si­tion and struc­ture of nat­ural micro­bial ecosys­tems. So it is chal­leng­ing for ecol­o­gists to get deep insights on how ecosys­tems func­tion and inter­play with sur­round­ing envi­ron­ments. But the recent pro­gresses in syn­thetic biol­ogy show that con­struc­tion of arti­fi­cial ecosys­tems where rela­tion­ships of species are com­par­a­tively clear could help us fur­ther uncover the meadow of those tiny soci­eties. By using two quorum-​​sensing sig­nal trans­duc­tion cir­cuits, this research designed, sim­u­lated and con­structed a syn­thetic ecosys­tem where var­i­ous pop­u­la­tion dynam­ics formed by chang­ing envi­ron­men­tal factors.”

“While some of the anti-​​fraud reme­dies of the secu­ri­ties laws still apply in 144A trans­ac­tions, these have been watered down in recent years by Con­gres­sional action and judi­cial inter­pre­ta­tion. In a series of opin­ions authored first by Jus­tice Pow­ell and then by Jus­tice Kennedy, the Supreme Court has steadily scaled back the scope of the secu­ri­ties laws. Opin­ions by Jus­tice Kennedy, in par­tic­u­lar, lim­ited the impact of anti-​​fraud pro­tec­tions as well as the abil­ity of investors to sue gate­keep­ers who play a sig­nif­i­cant role in prepar­ing offerings.”

“Now folks, it’s pretty reveal­ing when lob­by­ists have become so accus­tomed to their priv­i­leged access and back­room deal­ings with politi­cians –as went on in regards to Cheney’s energy dis­cus­sions, and each of the Bush tax cuts drawn up by a secre­tive group of GOP with­out any sun­light (or bipar­ti­san­sip), for exam­ple, and too much with the health care bill as well–that they don’t even bother to hide their scorn for the public’s views and their hopes for get­ting that back room deal to go their way. No won­der Wall Street hon­chos have been so brazenly arro­gant about their “enti­tle­ment” to bonuses, their rights to con­tinue pro­pri­etary trad­ing and hedge funds and deriv­a­tives desks–“doing God’s work” says Gold­man CEO Blankfein–when they are merely run­ning a casino mar­ket to strip as much gold off suck­ers as pos­si­ble with their “finan­cial inno­va­tions” like syn­thetic CDOs that made the mar­ket many times more volatile than “real” securitizations…”

“Since the micro­scope has no lenses, it does not mag­nify the images. Yet it is able to gain res­o­lu­tion just under two microm­e­ters, and makes images that are about as clear as those made by a con­ven­tional 40X micro­scope. This is made pos­si­ble by image-​​processing soft­ware. “We com­pen­sate for every­thing in the dig­i­tal regime,” says Ozcan. As light from the LED passes through a given type of cell, the light bends or dif­fracts in a char­ac­ter­is­tic way depend­ing on the cell’s size, shape, and refrac­tive index. Data picked up by the light-​​sensing chip is car­ried to a cell phone for analy­sis. Ozcan has pre­vi­ously demon­strated run­ning soft­ware on the phone that con­sults a library of dif­frac­tion sig­na­tures char­ac­ter­is­tic of par­tic­u­lar cell types and bac­te­ria to iden­tify and count the cells in the sample.”

“A recent study by researchers in Yale and pub­lished in PNAS shows that there are sig­nif­i­cant dif­fer­ences between the net­work topolo­gies of liv­ing sys­tems like E. coli and com­plex pieces of soft­ware such as the Linux Oper­at­ing System.”