Tag Archives: data-analysis

Items of some interest:

Posted on by

These are my recent Pin​board​.in links:

  • [1006.5366] “Not only defended but also applied”: The per­ceived absur­dity of Bayesian inference

    “The mis­sion­ary zeal of many Bayesians of old has been matched, in the other direc­tion, by a view among some the­o­reti­cians that Bayesian meth­ods are absurd-​​not merely mis­guided but obvi­ously wrong in prin­ci­ple. We con­sider sev­eral exam­ples, begin­ning with Feller’s clas­sic text on prob­a­bil­ity the­ory and con­tin­u­ing with more recent cases such as the per­ceived Bayesian nature of the so-​​called dooms­day argu­ment. We ana­lyze in this note the intel­lec­tual back­ground behind var­i­ous mis­con­cep­tions about Bayesian sta­tis­tics, with­out aim­ing at a com­plete his­tor­i­cal cov­er­age of the rea­sons for this dismissal.”

    social-​​dynamics sta­tis­tics martial-​​arts-​​schools

  • [1206.3268] Fea­ture Selec­tion via Block-​​Regularized Regression

    “In this paper, we con­sid­ered the prob­lem of find­ing a sub­set of covari­ates in a high-​​dimensional space that affect the out­put vari­able when there is a block struc– ture in the covari­ates. In the con­text of asso­ci­a­tion map­ping, we pro­posed a regression-​​based model with a Markov chain prior that encodes the infor­ma­tion in the cor­re­la­tion struc­ture such as dis­tance and re– com­bi­na­tion rate between adja­cent SNP mark­ers. We demon­strated on the sim­u­lated and mouse data that our pro­posed algo­rithm can be used to iden­tify groups of SNP mark­ers as a rel­e­vant block of causal SNPs. The idea of rep­re­sent­ing the cor­re­la­tion struc­ture as a Markov chain in a vari­able selec­tion method to learn grouped rel­e­vant vari­ables can be gen­er­al­ized to use a graph­i­cal model as a prior in a vari­able selec­tion prob– lem to rep­re­sent an arbi­trary cor­re­la­tion struc­ture in vari­ables in a high-​​dimensional space. Another inter– est­ing exten­sion of the model is to model a struc­ture in out­put vari­ables as well when mea­sure­ments of mul– tiple out­put vari­ables are available.”

    sta­tis­tics bioin­for­mat­ics algo­rithms data-​​mining feature-​​extraction

  • Fil­ipe Kiss : A bet­ter git log

    “So, are you tired of this old and bored git log screen?”

    yes software-​​development git tricks-​​n-​​tips bash

  • Neu­roskep­tic: Brains are Dif­fer­ent on Macs

    “The paper goes into lots more detail, but the les­son for researchers is extremely sim­ple: don’t cross the streams of data-​​analysis. Set up your analy­sis stream and then use it on all of your data. Same hard­ware, same soft­ware, same set­tings. Imag­ine you’re doing a study com­par­ing brain struc­ture in two groups. Halfway through ana­lyz­ing your data, you upgrade your MacOS. All of the brains you ana­lyze after that will be, say, 5% “big­ger”. That’ll cer­tainly make your data much nois­ier, and if you hap­pen to ana­lyze most of Group A before Group B, it’ll give you a false pos­i­tive find­ing. Some­times you just can’t avoid changes in hard­ware or soft­ware — IT techs have a habit of upgrad­ing things with­out ask­ing — but in these cases, you should run the same data under the old and the new regime to see if it’s mak­ing a dif­fer­ence. Finally, it would be wrong to blame FreeSurfer for this. I’d be sur­prised if they were any worse than the other soft­ware pack­ages. Mix­ing and match­ing ver­sions is some­thing that the FreeSurfer devel­op­ers specif­i­cally warn against. This paper shows why.”

    data-​​analysis repro­ducibil­ity technical-​​assumptions anomalies-​​are-​​where-​​you-​​find-​​them

  • Plug: What is infer­en­tial­ism? « Odontomachus’s Blog

    “I’ve been crit­i­cal of objects and the idea of ref­er­ence for a while now. To me sen­tences and propo­si­tions, by virtue of their role as “moves” in social inter­ac­tions, are likely to have pri­or­ity in a prop­erly objec­tive account of mean­ing. Many puta­tive objects (e.g. cor­po­ra­tions or muta­ble dig­i­tal doc­u­ments) bor­der on being fic­tional, gain­ing their object­hood only through what we say about them; and many refer­ring phrases seem to refer to dif­fer­ent things, depend­ing on what is being pred­i­cated. I think this opin­ion would make me what Pere­grin calls a “strong infer­en­tial­ist”. Even­tu­ally I hope that think­ing clearly about seman­tics ought to (among other things) help bring calm to the cur­rent mass hys­te­ria which is the Seman­tic Web and Linked Data, and help steer all of that energy expen­di­ture to improve its consequence.”

    prag­ma­tism indirect-​​links phi­los­o­phy talking-​​about-​​thinking-​​and-​​the-​​reverse

  • [1206.3552] A Clas­si­fi­ca­tion for Com­mu­nity Dis­cov­ery Meth­ods in Com­plex Networks

    “In the last few years many real-​​world net­works have been found to show a so-​​called com­mu­nity struc­ture orga­ni­za­tion. Much effort has been devoted in the lit­er­a­ture to develop meth­ods and algo­rithms that can effi­ciently high­light this hid­den struc­ture of the net­work, tra­di­tion­ally by par­ti­tion­ing the graph. Since net­work rep­re­sen­ta­tion can be very com­plex and can con­tain dif­fer­ent vari­ants in the tra­di­tional graph model, each algo­rithm in the lit­er­a­ture focuses on some of these prop­er­ties and estab­lishes, explic­itly or implic­itly, its own def­i­n­i­tion of com­mu­nity. Accord­ing to this def­i­n­i­tion it then extracts the com­mu­ni­ties that are able to reflect only some of the fea­tures of real com­mu­ni­ties. The aim of this sur­vey is to pro­vide a man­ual for the com­mu­nity dis­cov­ery prob­lem. Given a meta def­i­n­i­tion of what a com­mu­nity in a social net­work is, our aim is to orga­nize the main cat­e­gories of com­mu­nity dis­cov­ery based on their own def­i­n­i­tion of com­mu­nity. Given a desired def­i­n­i­tion of com­mu­nity and the fea­tures of a prob­lem (size of net­work, direc­tion of edges, mul­ti­di­men­sion­al­ity, and so on) this review paper is designed to pro­vide a set of approaches that researchers could focus on.”

    via:cshalizi graph-​​theory com­mu­nity clas­si­fi­ca­tion algo­rithms nudge

  • [1205.0792] Exact Wavelets on the Ball

    “We develop an exact wavelet trans­form on the three-​​dimensional ball (i.e. on the solid sphere), which we name the fla­glet trans­form. For this pur­pose we first con­struct an exact har­monic trans­form on the radial line using damped Laguerre poly­no­mi­als and develop a cor­re­spond­ing quad­ra­ture rule. Com­bined with the spher­i­cal har­monic trans­form, this approach leads to a sam­pling the­o­rem on the ball and a novel three-​​dimensional decom­po­si­tion which we call the Fourier-​​Laguerre trans­form. We relate this new trans­form to the well-​​known Fourier-​​Bessel decom­po­si­tion and show that band-​​limitness in the Fourier-​​Laguerre basis is a suf­fi­cient con­di­tion to com­pute the Fourier-​​Bessel decom­po­si­tion exactly. We then con­struct the fla­glet trans­form on the ball through a har­monic tiling, which is exact thanks to the exact­ness of the Fourier-​​Laguerre trans­form (from which the name fla­glets is coined). The cor­re­spond­ing wavelet ker­nels have com­pact local­i­sa­tion prop­er­ties in real and har­monic space and their angu­lar aper­ture is invari­ant under radial trans­la­tion. We intro­duce a mul­tires­o­lu­tion algo­rithm to per­form the fla­glet trans­form rapidly, while cap­tur­ing all infor­ma­tion at each wavelet scale in the min­i­mal num­ber of sam­ples on the ball. Our imple­men­ta­tion of these new tools achieves float­ing point pre­ci­sion and is made pub­licly avail­able. We per­form numer­i­cal exper­i­ments demon­strat­ing the speed and accu­racy of these libraries and illus­trate their capa­bil­i­ties on a sim­ple denois­ing example.”

    wavelets geom­e­try representation-​​theory signal-​​processing answer-​​languages

  • [1205.3077] Efficiency-​​Revenue Trade-​​offs in Auctions

    “When agents with inde­pen­dent pri­ors bid for a sin­gle item, Myerson’s opti­mal auc­tion max­i­mizes expected rev­enue, whereas Vickrey’s second-​​price auc­tion opti­mizes social wel­fare. We address the nat­ural ques­tion of trade-​​offs between the two cri­te­ria, that is, auc­tions that opti­mize, say, rev­enue under the con­straint that the wel­fare is above a given level. If one allows for ran­dom­ized mech­a­nisms, it is easy to see that there are polynomial-​​time mech­a­nisms that achieve any point in the trade-​​off (the Pareto curve) between rev­enue and wel­fare. We inves­ti­gate whether one can achieve the same guar­an­tees using deter­min­is­tic mech­a­nisms. We pro­vide a neg­a­tive answer to this ques­tion by show­ing that this is a (weakly) NP-​​hard prob­lem. On the pos­i­tive side, we pro­vide polynomial-​​time deter­min­is­tic mech­a­nisms that approx­i­mate with arbi­trary pre­ci­sion any point of the trade-​​off between these two fun­da­men­tal objec­tives for the case of two bid­ders, even when the val­u­a­tions are cor­re­lated arbi­trar­ily. The major prob­lem left open by our work is whether there is such an algo­rithm for three or more bid­ders with inde­pen­dent val­u­a­tion distributions.”

    algo­rithms Pareto-​​front performance-​​measure multiobjective-​​optimization

  • Sym­bol­set

    “Sym­bol­sets are seman­tic sym­bol fonts. They work in mod­ern browsers and any­where Open­Type fea­tures are supported.”

    typog­ra­phy uni­code

  • [1204.6653] Elim­i­na­tion of Glass Arti­facts and Object Segmentation

    “Many images nowa­days are cap­tured from behind the glasses and may have cer­tain stains dis­crep­ancy because of glass and must be processed to make dif­fer­en­ti­a­tion between the glass and objects behind it. This research paper pro­poses an algo­rithm to remove the dam­aged or cor­rupted part of the image and make it con­sis­tent with other part of the image and to seg­ment objects behind the glass. The dam­aged part is removed using total vari­a­tion inpaint­ing method and seg­men­ta­tion is done using kmeans clus­ter­ing, anisotropic dif­fu­sion and water­shed trans­for­ma­tion. The final out­put is obtained by inter­po­la­tion. This algo­rithm can be use­ful to appli­ca­tions in which some part of the images are cor­rupted due to data trans­mis­sion or needs to seg­ment objects from an image for fur­ther processing.”

    image-​​segmentation image-​​processing nudge-​​targets algo­rithms

  • The whole of the law — Things from your life

    “But it’ll be your deci­sion, not iner­tia or fate. The ongo­ing cadence of ask­ing these ques­tions (and, maybe, the con­tent of any answers you come up with) will con­vene an open space for you to live in. A world where what­ever you do is right.”

    this

  • The Pirate Uni­ver­sity | Pirate university

    “The Pirate Uni­ver­sity is an on-​​line bul­letin board on which stu­dents post requests for aca­d­e­mic pub­li­ca­tions. You can com­pare it to an aca­d­e­mic wish list. Oth­ers, who know where to find these pub­li­ca­tions, reply and if pos­si­ble, pro­vide links to the resources searched. The Pirate Uni­ver­sity is not pro­vid­ing, stor­ing or shar­ing copy­righted mate­r­ial. An impor­tant ques­tion is if the upload­ing of arti­cles, pub­li­ca­tions is legal. If you are the copy­right holder of the arti­cle requested, there should be no prob­lem. Also in cer­tain cases, if you or your insti­tute have acquired the rights of the pub­li­ca­tion, or if it is free of rights, there shouldn’t be a prob­lem. It is prob­a­bly best to con­sult with your librar­ian to see which kind of pub­li­ca­tion is okay to share on the Internet.”

    academic-​​culture pub­lish­ing col­lab­o­ra­tion crowd­sourc­ing librar­i­ans open-​​access schol­ar­ship

  • [1206.3793] A dis­trib­uted classification/​estimation algo­rithm for sen­sor networks

    “…We pro­pose a novel coop­er­a­tive iter­a­tive algo­rithm which copes with the com­mu­ni­ca­tion con­straints imposed by the net­work and shows remark­able per­for­mance. Our main result is a rig­or­ous proof of the con­ver­gence of the algo­rithm and a char­ac­ter­i­za­tion of the limit behav­ior. We also show that, in the limit when the num­ber of sen­sors goes to infin­ity, the com­mon unknown para­me­ter is esti­mated with arbi­trary small error, while the clas­si­fi­ca­tion error con­verges to that of the opti­mal cen­tral­ized max­i­mum like­li­hood esti­ma­tor. We also show numer­i­cal results that val­i­date the the­o­ret­i­cal analy­sis and sup­port their pos­si­ble gen­er­al­iza­tion. We com­pare our strat­egy with the Expectation-​​Maximization algo­rithm and we dis­cuss trade-​​offs in terms of robust­ness, speed of con­ver­gence and imple­men­ta­tion simplicity.”

    distributed-​​processing collective-​​behavior sensor-​​networks algo­rithms nudge-​​targets

  • [1204.6391] Extend­ing par­tial rep­re­sen­ta­tions of func­tion graphs and per­mu­ta­tion graphs

    “Func­tion graphs are graphs rep­re­sentable by inter­sec­tions of con­tin­u­ous real-​​valued func­tions on the inter­val [0,1] and are known to be exactly the com­ple­ments of com­pa­ra­bil­ity graphs. As such they are rec­og­niz­able in poly­no­mial time. Func­tion graphs gen­er­al­ize per­mu­ta­tion graphs, which arise when all func­tions con­sid­ered are lin­ear. We focus on the prob­lem of extend­ing par­tial rep­re­sen­ta­tions, which gen­er­al­izes the recog­ni­tion prob­lem. We observe that for per­mu­ta­tion graphs an easy exten­sion of Golumbic’s com­pa­ra­bil­ity graph recog­ni­tion algo­rithm can be exploited. This approach fails for func­tion graphs. Nev­er­the­less, we present a polynomial-​​time algo­rithm for extend­ing a par­tial rep­re­sen­ta­tion of a graph by func­tions defined on the entire inter­val [0,1] pro­vided for some of the ver­tices. On the other hand, we show that if a par­tial rep­re­sen­ta­tion con­sists of func­tions defined on subin­ter­vals of [0,1], then the prob­lem of extend­ing this rep­re­sen­ta­tion to func­tions on the entire inter­val [0,1] becomes NP-​​complete.”

    graph-​​theory math-i-didn’t-know representation-​​theory ontol­ogy inter­est­ing

  • [1206.3294] Flex­i­ble Pri­ors for Exemplar-​​based Clustering

    “Exemplar-​​based clus­ter­ing meth­ods have been shown to pro­duce state-​​of-​​the-​​art results on a num­ber of syn­thetic and real-​​world clus­ter­ing prob­lems. They are appeal­ing because they offer com­pu­ta­tional ben­e­fits over latent-​​mean mod­els and can han­dle arbi­trary pair­wise sim­i­lar­ity mea­sures between data points. How­ever, when try­ing to recover under­ly­ing struc­ture in clus­ter­ing prob­lems, tai­lored sim­i­lar­ity mea­sures are often not enough; we also desire con­trol over the dis­tri­b­u­tion of clus­ter sizes. Pri­ors such as Dirich­let process pri­ors allow the num­ber of clus­ters to be unspec­i­fied while express­ing pri­ors over data par­ti­tions. To our knowl­edge, they have not been applied to exemplar-​​based mod­els. We show how to incor­po­rate pri­ors, includ­ing Dirich­let process pri­ors, into the recently intro­duced affin­ity prop­a­ga­tion algo­rithm. We develop an effi­cient max­prod­uct belief prop­a­ga­tion algo­rithm for our new model and demon­strate exper­i­men­tally how the expanded range of clus­ter­ing pri­ors allows us to bet­ter recover true clus­ter­ings in sit­u­a­tions where we have some infor­ma­tion about the gen­er­at­ing process.”

    clus­ter­ing algo­rithms

  • Mag­a­zine — The Case Against Cre­den­tial­ism — The Atlantic

    ’”ALL OF OUR WORK HAS GIVEN ME A VERY STRONG view,” Richard Boy­atzis told me one after­noon. The con­sult­ing firm Boy­atzis heads, McBer and Com­pany, was founded by David McClel­land in 1963. Its spe­cialty has been ana­lyz­ing what peo­ple actu­ally do in busi­ness jobs—not what their job descrip­tions say, but how they spend their time and which skills seem most impor­tant to their suc­cess. “I’ve come to see that when­ever a group insti­tutes a cre­den­tial­ing process, whether by licens­ing or insist­ing on advanced degrees, the espoused rhetoric is to enforce the stan­dards of pro­fes­sion­al­ism. This is true whether it’s among accoun­tants or plumbers or physi­cians. But the observed con­se­quences always seem to be these two: the exclu­sion of cer­tain groups, whether by inten­tion or not, and the estab­lish­ment of mediocre per­for­mance standards.“‘

    pro­fes­sion­al­iza­tion cre­den­tial­ing Andrew-​​Abbott-​​smiles-​​in-​​Chicago author­ity exper­tise cultural-​​assumptions disintermediation-​​targets

  • [1205.2483] Edge-​​clique graphs of cock­tail par­ties have unbounded rankwidth

    “In an attempt to find a polynomial-​​time algo­rithm for the edge-​​clique cover prob­lem on cographs we tried to prove that the edge-​​clique graphs of cographs have bounded rankwidth. How­ever, this is not the case. In this note we show that the edge-​​clique graphs of cock­tail party graphs have unbounded rank width.”

    open-​​questions nudge-​​targets graph-​​theory algo­rithms

  • [1206.3235] Iden­ti­fy­ing rea­son­ing pat­terns in games

    “We present an algo­rithm that iden­ti­fies the rea­son­ing pat­terns of agents in a game, by iter­a­tively exam­in­ing the graph struc­ture of its Multi-​​Agent Influ­ence Dia­gram (MAID) rep­re­sen­ta­tion. If the deci­sion of an agent par­tic­i­pates in no rea­son­ing pat­terns, then we can effec­tively ignore that deci­sion for the pur­pose of cal­cu­lat­ing a Nash equi­lib­rium for the game. In some cases, this can lead to expo­nen­tial time sav­ings in the process of equi­lib­rium cal­cu­la­tion. More­over, our algo­rithm can be used to enu­mer­ate the rea­son­ing pat­terns in a game, which can be use­ful for con­struct­ing more effec­tive com­put­er­ized agents inter­act­ing with humans.”

    game-​​theory infer­ence strat­egy nudge-​​targets learning-​​by-​​watching

Items of some interest:

Posted on by

These are my recent Pin​board​.in links:

Items of some interest:

Posted on by

These are my recent Pin​board​.in links:

  • Wel­come to the Group Pat­tern Lan­guage Project | Group Works

    “This deck of 91 full-​​colour cards names what skilled facil­i­ta­tors and other par­tic­i­pants do to make things work.  The con­tent is more spe­cific than val­ues and less spe­cific than tips and tech­niques, cut­ting across exist­ing method­olo­gies with a designer’s eye to cap­ture the pat­terns that repeat.  The deck can be used to plan sess­sions, reflect on and debrief them, pro­vide guid­ance, and share respon­si­bil­ity for mak­ing the process go well.  It has the poten­tial to pro­vide a com­mon ref­er­ence point for prac­ti­tion­ers, and serve as a frame­work and learn­ing tool for those study­ing the field. ”

    via:bkerr col­lab­o­ra­tion design-​​patterns tools social-​​dynamics

  • [1202.0001] Vector-​​based model of elas­tic bonds for DEM sim­u­la­tion of solids

    “A new model for com­puter sim­u­la­tion of solids, com­posed of bonded par­ti­cles, is pro­posed. Vec­tors rigidly con­nected with par­ti­cles are used for descrip­tion of defor­ma­tion of a sin­gle bond. The expres­sion for poten­tial energy of the bond and cor­re­spond­ing expres­sions for forces and moments are pro­posed. For­mu­las, con­nect­ing para­me­ters of the model with lon­gi­tu­di­nal, shear, bend­ing and tor­sional stiff­nesses of the bond, are derived. It is shown that the model allows to describe any val­ues of the bond stiff­nesses exactly. Two dif­fer­ent cal­i­bra­tion pro­ce­dures depend­ing on bond length/​thickness ratio are pro­posed. It is shown that para­me­ters of model can be cho­sen so that under small defor­ma­tions the bond is equiv­a­lent to either Bernoulli-​​Euler or Tim­o­shenko rod or short cylin­der con­nect­ing par­ti­cles. Sim­ple expres­sions, con­nect­ing para­me­ters of V-​​model with geo­met­ri­cal and mechan­i­cal char­ac­ter­is­tics of the bond, are derived. Com­puter sim­u­la­tion of dynam­i­cal buck­ling of the straight dis­crete rod and dis­crete half-​​spherical shell is car­ried out.”

    mod­el­ing mechanical-​​systems materials-​​science computational-​​methods algo­rithms nudge-​​targets

  • [1202.0253] High-​​speed Flight in an Ergodic Forest

    “Inspired by birds fly­ing through clut­tered envi­ron­ments such as dense forests, this paper stud­ies the the­o­ret­i­cal foun­da­tions of a novel motion plan­ning prob­lem: high-​​speed nav­i­ga­tion through a randomly-​​generated obsta­cle field when only the sta­tis­tics of the obsta­cle gen­er­at­ing process are known a pri­ori. Resem­bling a pla­nar for­est envi­ron­ment, the obsta­cle gen­er­at­ing process is assumed to deter­mine the loca­tions and sizes of disk-​​shaped obsta­cles. When this process is ergodic, and under mild tech­ni­cal con­di­tions on the dynam­ics of the bird, it is shown that the exis­tence of an infi­nite collision-​​free tra­jec­tory through the for­est exhibits a phase tran­si­tion. On one hand, if the bird flies faster than a cer­tain crit­i­cal speed, then, with prob­a­bil­ity one, there is no infi­nite collision-​​free tra­jec­tory, i.e., the bird will even­tu­ally col­lide with some tree, almost surely, regard­less of the plan­ning algo­rithm gov­ern­ing the bird’s motion. On the other hand, if the bird flies slower than this crit­i­cal speed, then there exists at least one infi­nite collision-​​free tra­jec­tory, almost surely. Lower and upper bounds on the crit­i­cal speed are derived for the spe­cial case of a homo­ge­neous Pois­son for­est con­sid­er­ing a sim­ple model for the bird’s dynam­ics. For the same case, an equiv­a­lent per­co­la­tion model is pro­vided. Using this model, the phase dia­gram is approx­i­mated in Monte-​​Carlo sim­u­la­tions. This paper also estab­lishes novel con­nec­tions between robot motion plan­ning and sta­tis­ti­cal physics through ergodic the­ory and per­co­la­tion the­ory, which may be of inde­pen­dent interest.”

    robot­ics plan­ning algo­rithms nudge-​​targets

  • [1202.0077] An Inter­act­ing Par­ti­cle Model for Clus­ter­ing Euclid­ean Datasets

    “In this paper we pro­pose a method based on inter­act­ing par­ti­cle physics, devised for clus­ter­ing Euclid­ean datasets with­out ini­tial con­straints or con­di­tions. We model any dataset as an inter­act­ing par­ti­cle sys­tem, whose ele­ments cor­re­spond to par­ti­cles that inter­act through a sim­pli­fied ver­sion of Lennard-​​Jones poten­tials. In so doing, mutual attrac­tive inter­ac­tions allow to iden­tify groups of prox­i­mal par­ti­cles. The main out­come of this mod­el­ing task is an adja­cency matrix, taken as input by a com­mu­nity detec­tion algo­rithm aimed to iden­tify dif­fer­ent par­ti­tions. The under­ly­ing con­jec­ture is that, using a mul­tires­o­lu­tion analy­sis, the adopted model allows to find the right num­ber of clus­ters for any given dataset. Exper­i­men­tal results, per­formed in com­par­i­son with a clas­si­cal clus­ter­ing algo­rithm, con­firm this assumption.”

    clus­ter­ing data-​​analysis algo­rithms nudge-​​targets distributed-​​processing

  • [1201.6583] Empow­er­ment for Con­tin­u­ous Agent-​​Environment Systems

    “This paper devel­ops gen­er­al­iza­tions of empow­er­ment to con­tin­u­ous states. Empow­er­ment is a recently intro­duced information-​​theoretic quan­tity moti­vated by hypothe­ses about the effi­ciency of the sen­so­ri­mo­tor loop in bio­log­i­cal organ­isms, but also from con­sid­er­a­tions stem­ming from curiosity-​​driven learn­ing. Empowe­mer­ment mea­sures, for agent-​​environment sys­tems with sto­chas­tic tran­si­tions, how much influ­ence an agent has on its envi­ron­ment, but only that influ­ence that can be sensed by the agent sen­sors. It is an information-​​theoretic gen­er­al­iza­tion of joint con­trol­la­bil­ity (influ­ence on envi­ron­ment) and observ­abil­ity (mea­sure­ment by sen­sors) of the envi­ron­ment by the agent, both con­trol­la­bil­ity and observ­abil­ity being usu­ally defined in con­trol the­ory as the dimen­sion­al­ity of the control/​observation spaces.…”

    agent-​​based emergent-​​design robot­ics engineering-​​design machine-​​learning empow­er­ment nudge

  • [1201.6655] Learn­ing Per­for­mance of Pre­dic­tion Mar­kets with Kelly Bettors

    “In eval­u­at­ing pre­dic­tion mar­kets (and other crowd-​​prediction mech­a­nisms), inves­ti­ga­tors have repeat­edly observed a so-​​called “wis­dom of crowds” effect, which roughly says that the aver­age of par­tic­i­pants per­forms much bet­ter than the aver­age par­tic­i­pant. The mar­ket price—an aver­age or at least aggre­gate of traders’ beliefs—offers a bet­ter esti­mate than most any indi­vid­ual trader’s opin­ion. In this paper, we ask a stronger ques­tion: how does the mar­ket price com­pare to the best trader’s belief, not just the aver­age trader. We mea­sure the market’s worst-​​case log regret, a notion com­mon in machine learn­ing the­ory. To arrive at a mean­ing­ful answer, we need to assume some­thing about how traders behave. We sup­pose that every trader opti­mizes accord­ing to the Kelly cri­te­ria, a strat­egy that prov­ably max­i­mizes the com­pound growth of wealth over an (infi­nite) sequence of mar­ket inter­ac­tions. We show sev­eral consequences.…”

    pre­dic­tion performance-​​measure agent-​​based sim­u­la­tion nudge-​​targets wisdom-​​of-​​crowds

  • Curat­ing the kraken « Pub­lic Historian

    ‘This is why “curate” is still a word to con­jure by in our cul­ture.  It still promises trans­for­ma­tive power.’

    muse­ol­ogy prag­mat­ics nam­ing engineering-​​of-​​philosophy

  • [1201.5780] Full and Half Gilbert Tes­sel­la­tions with Rec­tan­gu­lar Cells

    “We inves­ti­gate the ray-​​length dis­tri­b­u­tions for two dif­fer­ent rec­tan­gu­lar ver­sions of Gilbert’s tes­sel­la­tion. In the full rec­tan­gu­lar ver­sion, lines extend either hor­i­zon­tally (with east– and west-​​growing rays) or ver­ti­cally (north– and south-​​growing rays) from seed points which form a Pois­son point process, each ray stop­ping when another ray is met. In the half rec­tan­gu­lar ver­sion, east and south grow­ing rays do not inter­act with west and north rays. For the half rec­tan­gu­lar tes­sel­la­tion we com­pute ana­lyt­i­cally, via recur­sion, a series expan­sion for the ray-​​length dis­tri­b­u­tion, whilst for the full rec­tan­gu­lar ver­sion we develop an accu­rate sim­u­la­tion tech­nique, based in part on the stopping-​​set the­ory of Zuyev, to accom­plish the same. We demon­strate the remark­able fact that plots of the two dis­tri­b­u­tions appear to be iden­ti­cal when the inten­sity of seeds in the half model is twice that in the full model. Our paper explores this coin­ci­dence mind­ful of the fact that, for one model, our results are from a sim­u­la­tion (with inher­ent sam­pling error).…”

    geom­e­try tiling algo­rithms generative-​​art sim­u­la­tion emer­gence interesting-​​problem

Items of some interest:

Posted on by

These are my recent Pin​board​.in links:

  • [1201.5604] Dis­crete and Fuzzy Dynam­i­cal Genetic Pro­gram­ming in the XCSF Learn­ing Clas­si­fier System

    “A num­ber of rep­re­sen­ta­tion schemes have been pre­sented for use within Learn­ing Clas­si­fier Sys­tems, rang­ing from binary encod­ings to neural net­works. This paper presents results from an inves­ti­ga­tion into using dis­crete and fuzzy dynam­i­cal sys­tem rep­re­sen­ta­tions within the XCSF Learn­ing Clas­si­fier Sys­tem. In par­tic­u­lar, asyn­chro­nous Ran­dom Boolean Net­works are used to rep­re­sent the tra­di­tional condition-​​action pro­duc­tion sys­tem rules in the dis­crete case and asyn­chro­nous Fuzzy Logic Net­works in the continuous-​​valued case. It is shown pos­si­ble to use self-​​adaptive, open-​​ended evo­lu­tion to design an ensem­ble of such dynam­i­cal sys­tems within XCSF to solve a num­ber of well-​​known test problems.”

    Kauffman-​​networks learning-​​classifier-​​systems genetic-​​programming nudge-​​targets inter­est­ing

  • [1201.4899] I Like Her more than You: Self-​​determined Communities

    “In this paper we define what we call an affin­ity sys­tem, which is a set of indi­vid­u­als, each with a vec­tor char­ac­ter­iz­ing its pref­er­ence for all other indi­vid­u­als in the set. The pref­er­ence of a mem­ber can be given either by a rank­ing of all mem­bers or by a weighted vec­tor that defines the degrees of its affin­ity to oth­ers. Affin­ity sys­tems are use­ful for mod­el­ing social sys­tems as well as gen­eral data sets, as social inter­ac­tions are often deter­mined by affini­ties among the mem­bers. We also define a nat­ural notion of (poten­tially over­lap­ping) com­mu­ni­ties in an affin­ity sys­tem, in which the mem­bers of a given com­mu­nity col­lec­tively pre­fer each other to any­one else out­side the com­mu­nity. Thus these com­mu­ni­ties are “self-​​determined” or “self-​​certified” by the affin­ity sys­tem. We pro­vide a tight poly­no­mial bound on the num­ber of self-​​determined com­mu­ni­ties as a func­tion of the robust­ness of the com­mu­nity. More­over, we present a polynomial-​​time algo­rithm for enu­mer­at­ing these com­mu­ni­ties, as well as a local algo­rithm with a strong sto­chas­tic per­for­mance guar­an­tee that can find a com­mu­nity in time nearly lin­ear in the of size the community.…”

    network-​​theory social-​​capital social-​​dynamics self-​​assembly agent-​​based graph-​​theory algo­rithms com­plex­ol­ogy nudge-​​targets

  • [1201.5076] Tech­ni­cal Report #SEHIR-IE-VA-12–1: Opti­mal Obsta­cle Place­ment with Disambiguations

    “We intro­duce the opti­mal obsta­cle place­ment with dis­am­bigua­tions prob­lem wherein the goal is to place true obsta­cles in an envi­ron­ment clut­tered with false obsta­cles so as to max­i­mize the total tra­ver­sal length of a nav­i­gat­ing agent (NAVA). Prior to the tra­ver­sal, NAVA is given loca­tion infor­ma­tion and prob­a­bilis­tic esti­mates of each disk-​​shaped hin­drance (here­inafter referred to as disk) being a true obsta­cle. The NAVA can dis­am­biguate a disk’s sta­tus only when sit­u­ated on its bound­ary. There exists an obsta­cle plac­ing agent (OPA) that locates obsta­cles prior to NAVA’s tra­ver­sal. The goal of OPA is to place true obsta­cles in between the clut­ter in such a way that NAVA’s tra­ver­sal length is max­i­mized in a game-​​theoretic sense.…”

    agent-​​based game-​​theory robot­ics disambiguation-​​design nudge-​​targets military-​​applications algo­rithms

  • [1010.5017] Col­lec­tive motion

    “We review the obser­va­tions and the basic laws describ­ing the essen­tial aspects of col­lec­tive motion — being one of the most com­mon and spec­tac­u­lar man­i­fes­ta­tion of coor­di­nated behav­ior. Our aim is to pro­vide a bal­anced dis­cus­sion of the var­i­ous facets of this highly mul­ti­dis­ci­pli­nary field, includ­ing exper­i­ments, math­e­mat­i­cal meth­ods and mod­els for sim­u­la­tions, so that read­ers with a vari­ety of back­ground could get both the basics and a broader, more detailed pic­ture of the field. The obser­va­tions we report on include sys­tems con­sist­ing of units rang­ing from macro­mol­e­cules through metal­lic rods and robots to groups of ani­mals and peo­ple. Some empha­sis is put on mod­els that are sim­ple and real­is­tic enough to repro­duce the numer­ous related obser­va­tions and are use­ful for devel­op­ing con­cepts for a bet­ter under­stand­ing of the com­plex­ity of sys­tems con­sist­ing of many simul­ta­ne­ously mov­ing enti­ties. As such, these mod­els allow the estab­lish­ing of a few fun­da­men­tal prin­ci­ples of flock­ing. In par­tic­u­lar, it is demon­strated, that in spite of con­sid­er­able dif­fer­ences, a num­ber of deep analo­gies exist between equi­lib­rium sta­tis­ti­cal physics sys­tems and those made of self-​​propelled (in most cases liv­ing) units. In both cases only a few well defined macroscopic/​collective states occur and the tran­si­tions between these states fol­low a sim­i­lar sce­nario, involv­ing dis­con­ti­nu­ity and alge­braic divergences.”

    emer­gence emergent-​​design biol­ogy ethol­ogy com­plex­ol­ogy mod­els artificial-​​life nudge-​​targets

  • [1201.5568] Dynamic trees for stream­ing and mas­sive data contexts

    “Data col­lec­tion at a mas­sive scale is becom­ing ubiq­ui­tous in a wide vari­ety of set­tings, from vast offline data­bases to stream­ing real-​​time infor­ma­tion. Learn­ing algo­rithms deployed in such con­texts must rely on single-​​pass infer­ence, where the data his­tory is never revis­ited. In stream­ing con­texts, learn­ing must also be tem­po­rally adap­tive to remain up-​​to-​​date against unfore­seen changes in the data gen­er­at­ing mech­a­nism. Although rapidly grow­ing, the online Bayesian infer­ence lit­er­a­ture remains chal­lenged by mas­sive data and tran­sient, evolv­ing data streams. Non-​​parametric mod­el­ling tech­niques can prove par­tic­u­larly ill-​​suited, as the com­plex­ity of the model is allowed to increase with the sam­ple size. In this work, we take steps to over­come these chal­lenges by port­ing stan­dard stream­ing tech­niques, like data dis­card­ing and down­weight­ing, into a fully Bayesian frame­work via the use of infor­ma­tive pri­ors and active learn­ing heuris­tics. We show­case our meth­ods by aug­ment­ing a mod­ern non-​​parametric mod­el­ling frame­work, dynamic trees, and illus­trate its per­for­mance on a num­ber of prac­ti­cal exam­ples. The end prod­uct is a pow­er­ful stream­ing regres­sion and clas­si­fi­ca­tion tool, whose per­for­mance com­pares favourably to the state-​​of-​​the-​​art.”

    data-​​analysis learning-​​from-​​data algo­rithms drinking-​​from-​​the-​​firehose nudge data-​​mining

Items of some interest…

Posted on by

These are my recent Pin​board​.in links:

  • Pro­gres­sives and the Ron Paul fal­lac­ies — Salon​.com

    The fal­lacy in this rea­son­ing is glar­ing. The can­di­date sup­ported by pro­gres­sives — Pres­i­dent Obama — him­self holds heinous views on a slew of crit­i­cal issues and him­self has done heinous things with the power he has been vested. He has slaugh­tered civil­ians — Mus­lim chil­dren by the dozens — not once or twice, but con­tin­u­ously in numer­ous nations with drones, cluster bombs and other forms of attack. He has sought to over­turn a global ban on clus­ter bombs. He has insti­tu­tion­al­ized the power of Pres­i­dents — in secret and with no checks — to tar­get Amer­i­can cit­i­zens for assassination-​​by-​​CIA, far from any bat­tle­field. He has waged an unprece­dented war against whistle­blow­ers, the pro­tec­tion of which was once a lib­eral shib­bo­leth. He ren­dered per­ma­nently irrel­e­vant the War Pow­ers Res­o­lu­tion, a crown jewel in the list of post-​​Vietnam lib­eral accom­plish­ments, and thus enshrined the power of Pres­i­dents to wage war even in the face of a Con­gres­sional vote against it. His obses­sion with secrecy is so extreme that it has become darkly laugh­able in its man­i­fes­ta­tions, and he even worked to amend the Free­dom of Infor­ma­tion Act (another crown jewel of lib­eral leg­isla­tive suc­cesses) when com­pli­ance became inconvenient.

    pol­i­tics party-​​politics-​​in-​​particular cognitive-​​dissonance cultural-​​assumptions dialog-it-ain’t

  • A mod­est pro­posal to give Free Soft­ware equal legal stand­ing as pro­pri­etary. | Carlo Piana :: Law is Freedom ::

    Laws are more often than not an annoy­ance, despite their aim to improve the legal frame­work in any given field. Free Soft­ware (AKA “Open Source”) has thrieved despite the absence of any legal recog­ni­tion by the law, if not in spite of rules that clearly are shaped around pro­pri­etary soft­ware. In many juris­dic­tions it has passed the enforce­abil­ity test. So, no laws seem nec­es­sary to make it work. Yet, can some legal prin­ci­ple be put for­ward, and included in some laws, to help?

    via:Glyn-Moody licens­ing law con­tracts modest-​​proposals

  • 2011 in Review


    to-​​read to-​​keep-​​in-​​mind lists movies books comix

  • Best and Worst Sci­ence Fiction/​Fantasy Movies of 2011


    to-​​keep-​​in-​​mind movies lists

  • [1109.3248] Recon­struc­tion of sequen­tial data with den­sity models

    We intro­duce the prob­lem of recon­struct­ing a sequence of mul­ti­di­men­sional real vec­tors where some of the data are miss­ing. This prob­lem con­tains regres­sion and map­ping inver­sion as par­tic­u­lar cases where the pat­tern of miss­ing data is inde­pen­dent of the sequence index. The prob­lem is hard because it involves pos­si­bly mul­ti­val­ued map­pings at each vec­tor in the sequence, where the miss­ing vari­ables can take more than one value given the present vari­ables; and the set of miss­ing vari­ables can vary from one vec­tor to the next. To solve this prob­lem, we pro­pose an algo­rithm based on two redun­dancy assump­tions: vec­tor redun­dancy (the data live in a low-​​dimensional man­i­fold), so that the present vari­ables con­strain the miss­ing ones; and sequence redun­dancy (e.g. con­ti­nu­ity), so that con­sec­u­tive vec­tors con­strain each other. We cap­ture the low-​​dimensional nature of the data in a prob­a­bilis­tic way with a joint den­sity model, here the gen­er­a­tive topo­graphic map­ping, which results in a Gauss­ian mix­ture. Can­di­date recon­struc­tions at each vec­tor are obtained as all the modes of the con­di­tional dis­tri­b­u­tion of miss­ing vari­ables given present vari­ables. The recon­structed sequence is obtained by min­imis­ing a global con­straint, here the sequence length, by dynamic pro­gram­ming. We present exper­i­men­tal results for a toy prob­lem and for inverse kine­mat­ics of a robot arm.

    inverse-​​problems sta­tis­tics algo­rithms learning-​​from-​​data nudge-​​targets

  • [1110.5063] Recov­er­ing a Clipped Sig­nal in Sparseland

    In many data acqui­si­tion sys­tems it is com­mon to observe sig­nals whose ampli­tudes have been clipped. We present two new algo­rithms for recov­er­ing a clipped sig­nal by lever­ag­ing the model assump­tion that the under­ly­ing sig­nal is sparse in the fre­quency domain. Both algo­rithms employ ideas com­monly used in the field of Com­pres­sive Sens­ing; the first is a mod­i­fied ver­sion of Reweighted $ell_​1$ min­i­miza­tion, and the sec­ond is a mod­i­fi­ca­tion of a sim­ple greedy algo­rithm known as Triv­ial Pur­suit. An empir­i­cal inves­ti­ga­tion shows that both approaches can recover sig­nals with sig­nif­i­cant lev­els of clipping

    signal-​​processing infer­ence compressive-​​sensing algo­rithms nudge-​​targets

  • [1112.2316] Complexity-​​entropy causal­ity plane: a use­ful approach for dis­tin­guish­ing songs

    Nowa­days we are often faced with huge data­bases result­ing from the rapid growth of data stor­age tech­nolo­gies. This is par­tic­u­larly true when deal­ing with music data­bases. In this con­text, it is essen­tial to have tech­niques and tools able to dis­crim­i­nate prop­er­ties from these mas­sive sets. In this work, we report on a sta­tis­ti­cal analy­sis of more than ten thou­sand songs aim­ing to obtain a com­plex­ity hier­ar­chy. Our approach is based on the esti­ma­tion of the per­mu­ta­tion entropy com­bined with an inten­sive com­plex­ity mea­sure, build­ing up the complexity-​​entropy causal­ity plane. The results obtained indi­cate that this rep­re­sen­ta­tion space is very promis­ing to dis­crim­i­nate songs as well as to allow a rel­a­tive quan­ti­ta­tive com­par­i­son among songs. Addi­tion­ally, we believe that the here-​​reported method may be applied in prac­ti­cal sit­u­a­tions since it is sim­ple, robust and has a fast numer­i­cal implementation.

    signal-​​processing clas­si­fi­ca­tion data-​​analysis clus­ter­ing rep­re­sen­ta­tion music nudge-​​targets

  • [1112.6178] A gen­eral frame­work for online audio source separation

    We con­sider the prob­lem of online audio source sep­a­ra­tion. Exist­ing algo­rithms adopt either a slid­ing block approach or a sto­chas­tic gra­di­ent approach, which is faster but less accu­rate. Also, they rely either on spa­tial cues or on spec­tral cues and can­not sep­a­rate cer­tain mix­tures. In this paper, we design a gen­eral online audio source sep­a­ra­tion frame­work that com­bines both approaches and both types of cues. The model para­me­ters are esti­mated in the Max­i­mum Like­li­hood (ML) sense using a Gen­er­alised Expec­ta­tion Max­imi­sa­tion (GEM) algo­rithm with mul­ti­plica­tive updates. The sep­a­ra­tion per­for­mance is eval­u­ated as a func­tion of the block size and the step size and com­pared to that of an offline algorithm.

    signal-​​processing audio-​​segmentation sta­tis­tics algo­rithms meta­heuris­tics nudge-​​targets