Items of some interest:

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

  • Attrac­tive Mod­els — Kieran Healy

    “Now, if you write a paper describ­ing neg­a­tive results—a model where noth­ing is significant—then you may have a hard time get­ting it pub­lished. In the absence of some spe­cific con­tro­versy, neg­a­tive results are bor­ing. For the same rea­son, though, if your results just barely cross the thresh­old of con­ven­tional sig­nif­i­cance, they may stand a dis­pro­por­tion­ately bet­ter chance of get­ting pub­lished than an oth­er­wise quite sim­i­lar paper where the results just failed to make the thresh­old. And this is what the graph above shows, for papers pub­lished in the Amer­i­can Polit­i­cal Sci­ence Review. It’s a his­togram of p-​​values for coef­fi­cients in regres­sions reported in the jour­nal. The dashed line is the con­ven­tional thresh­old for sig­nif­i­cance. The tall red bar to the right of the dashed line is the num­ber of coef­fi­cients that just made it over the thresh­old, while the short red bar is the num­ber of coef­fi­cients that just failed to do so. If there were no bias in the pub­li­ca­tion process, the shape of the his­togram would approx­i­mate the right-​​hand side of a bell curve. The gap between the big and the small red bars is a con­se­quence of two things: the unwill­ing­ness of jour­nals to report neg­a­tive results, and the efforts of authors to search for (and write up) results that cross the con­ven­tional threshold.”

    sta­tis­tics academic-​​culture pub­lish­ing meta-​​analysis

Items of some interest:

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

  • [1204.4366] Multipath-​​dominant, pulsed doppler analy­sis of rotat­ing blades

    “We present a novel angu­lar fin­ger­print­ing algo­rithm for detect­ing changes in the direc­tion of rota­tion of a tar­get with a mono­sta­tic, sta­tion­ary sonar plat­form. Unlike other approaches, we assume that the target’s cen­troid is sta­tion­ary, and exploit doppler mul­ti­path sig­nals to resolve the oth­er­wise unavoid­able ambi­gu­i­ties that arise. Since the algo­rithm is based on an under­ly­ing dif­fer­en­tial topo­log­i­cal the­ory, it is highly robust to dis­tor­tions in the col­lected data. We demon­strate per­for­mance of this algo­rithm exper­i­men­tally, by exhibit­ing a pulsed doppler sonar col­lec­tion sys­tem that runs on a smart­phone. The per­for­mance of this sys­tem is suf­fi­ciently good to both detect changes in tar­get rota­tion direc­tion using angu­lar fin­ger­prints, and also to form high-​​resolution inverse syn­thetic aper­a­ture images of the target.”

    signal-​​processing algo­rithms radar nudge-​​targets the-​​imperial-​​we
  • [1204.3850] Sim­ple Agents Learn to Find Their Way: An Intro­duc­tion on Map­ping Polygons

    “This paper gives an intro­duc­tion to the prob­lem of map­ping sim­ple poly­gons with autonomous agents. We focus on min­i­mal­is­tic agents that move from ver­tex to ver­tex along straight lines inside a poly­gon, using their sen­sors to gather local obser­va­tions at each ver­tex. Our atten­tion revolves around the ques­tion whether a given con­fig­u­ra­tion of sen­sors and move­ment capa­bil­i­ties of the agents allows them to cap­ture enough data in order to draw con­clu­sions regard­ing the global lay­out of the poly­gon. In par­tic­u­lar, we study the prob­lem of recon­struct­ing the vis­i­bil­ity graph of a sim­ple poly­gon by an agent mov­ing either inside or on the bound­ary of the poly­gon. Our aim is to pro­vide insight about the algo­rith­mic chal­lenges faced by an agent try­ing to map a poly­gon. We present an overview of tech­niques for solv­ing this prob­lem with agents that are equipped with sim­ple sen­so­r­ial capa­bil­i­ties. We illus­trate these tech­niques on exam­ples with sen­sors that mea– sure angles between lines of sight or iden­tify the pre­vi­ous loca­tion. We give an overview over related prob­lems in com­bi­na­to­r­ial geom­e­try as well as graph exploration.”

    agent-​​based algo­rithms nudge-​​targets
  • [1204.4202] Fuzzy Dynam­i­cal Genetic Pro­gram­ming in XCSF

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

    learning-​​classifier-​​systems genetic-​​programming fuzzy-​​math dynamical-​​control rules-​​learning nudge-​​targets
  • Omni­scient Gen­tle­men of The Atlantic | | Note­book | The Baffler

    “What mys­ti­fied Grove was the asser­tion, voiced by the econ­o­mist Alan Blinder and oth­ers, “that as long as ‘knowl­edge work’ stays in the U.S., it doesn’t mat­ter what hap­pens to fac­tory jobs.” This was not only inhu­mane, Grove declared; it was idiotic.”

    via:cshalizi cor­po­ratism pub­lish­ing social-​​engineering jour­nal­ism they-​​say-​​the-​​best-​​astroturf-​​has-​​no-​​color-​​at-​​all
  • [1204.3293] Effi­ciently decod­ing strings from their shingles

    “Deter­min­ing whether an unordered col­lec­tion of over­lap­ping sub­strings (called shin­gles) can be uniquely decoded into a con­sis­tent string is a prob­lem that lies within the foun­da­tion of a broad assort­ment of dis­ci­plines rang­ing from net­work­ing and infor­ma­tion the­ory through cryp­tog­ra­phy and even genetic engi­neer­ing and lin­guis­tics. We present three per­spec­tives on this prob­lem: a graph the­o­retic frame­work due to Pevzner, an automata the­o­retic approach from our pre­vi­ous work, and a new insight that yields a time-​​optimal stream­ing algo­rithm for deter­min­ing whether a string of $n$ char­ac­ters over the alpha­bet $Sigma$ can be uniquely decoded from its two-​​character shin­gles. Our algo­rithm achieves an over­all time com­plex­ity $Theta(n)$ and space com­plex­ity $O(|Sigma|)$. As an appli­ca­tion, we demon­strate how this algo­rithm can be extended to larger shin­gles for effi­cient string reconciliation.”

    strings algo­rithms computational-​​complexity nudge-​​targets
  • Script­ing News: It’s def­i­nitely a bubble

    “They’re turn­ing uni­ver­si­ties into incu­ba­tors. It’s hap­pen­ing at NYU and Har­vard, two schools I have some famil­iar­ity with. Prob­a­bly every­where else too, to some extent. But I’d guess these two schools are pretty lead­ing edge. Stan­ford has been there for a few generations.”

    bub­ble entrepreneurship-​​as-​​pathology startup-​​culture-​​must-​​die ayup

  • via:cshalizi love­craft humor also-​​the-​​whole-​​zine-​​blog-​​thing
  • CodeMir­ror

    “CodeMir­ror is a JavaScript library that can be used to cre­ate a rel­a­tively pleas­ant edi­tor inter­face for code-​​like con­tent ― com­puter pro­grams, HTML markup, and sim­i­lar. If a mode has been writ­ten for the lan­guage you are edit­ing, the code will be coloured, and the edi­tor will option­ally help you with indentation.”

    javascript edi­tor library toolkit bookphile

Items of some interest:

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

  • A List Apart: Arti­cles: Artis­tic Distance

    “While I’m sure that some­one will dis­agree, these sites have proven that very few “pro­fes­sion­als” have the abil­ity or courage to pro­vide a well-​​constructed analy­sis of some­one else’s work (whether or not the eval­u­a­tion was solicited). My opin­ion has noth­ing at all to do with either web­site, but rather with indus­try pro­fes­sion­als’ inabil­ity to chal­lenge, or fear of chal­leng­ing, the sta­tus quo. Far too often, hon­esty is met with ridicule, shame, or out­right rage from peo­ple hid­ing behind elec­tronic media. As a com­mu­nity, if our goal is to con­tinue rais­ing the bar for design, we need to get to a place where objec­tive dis­cus­sion is wel­comed, not scorned or drowned in obse­quious­ness. I would love to see dis­cus­sion of basic design move past the super­fi­cial trendi­ness of emerg­ing web technologies.”

    cri­tique col­lab­o­ra­tion advice graphic-​​design not-​​just
  • - How We Will Read: Laura Miller and Maud Newton

    LM: Lit­er­ary peo­ple, when they talk about books, tend to think of fic­tion first. But most peo­ple, when they think about books, are think­ing about non­fic­tion, which lends itself amaz­ingly well to some kind of enhanced e-​​book expe­ri­ence. As a piece of that, I’m skep­ti­cal of enhanc­ing fic­tion e-​​books. The essence of nar­ra­tive is this sense of causal­ity and mean­ing, and when you intro­duce a lot of arbi­trary or ran­dom branch­ing things into it, it actu­ally loses it’s core plea­sure. It’s a tricky issue.”

    pub­lish­ing ebooks read­ing edi­tor
  • Per­sonal Tech for the 17th Cen­tury — Suzanne Fis­cher — Tech­nol­ogy — The Atlantic

    “The university’s John Carter Brown Library has long held the “Roger Williams Mys­tery Book,” a book that pur­port­edly belonged to Roger Williams, the rad­i­cal reli­gious thinker and founder of Rhode Island. The book is miss­ing its title page and thus has lit­tle iden­ti­fy­ing infor­ma­tion (besides a sub­ti­tle, “An Essay Con­cern­ing the Rec­on­cil­ing of Dif­fer­ences among Chris­tians”) — but it’s cov­ered with exten­sive short­hand mar­gin­a­lia sus­pected to have been writ­ten by Williams him­self some­time in the mid 1600s. The stu­dents, who include his­tory and math majors, are using this semes­ter to deci­pher the writ­ing and to deter­mine whether or not the short­hand hand­writ­ing was Williams’s hand.”

    nanohis­tory mar­gin­a­lia early-​​modern puz­zles
  • atomo

    “atomo is a small, sim­ple, insanely flex­i­ble and expres­sive pro­gram­ming lan­guage. its design is inspired by Scheme (small, sim­ple core), Slate (mul­ti­ple dis­patch, key­words), Ruby (very DSL-​​friendly), and Erlang (message-​​passing con­cur­rency). it is writ­ten in and pig­gy­backs on the Haskell run­time, per­mit­ting access to all of its power (and libraries!) through a thin layer.”

    pro­gram­ming lan­guage
  • Jour­nal of Dig­i­tal Humanities

    “The Jour­nal of Dig­i­tal Human­i­ties is a com­pre­hen­sive, peer-​​reviewed, open access jour­nal that fea­tures the best schol­ar­ship, tools, and con­ver­sa­tions pro­duced by the dig­i­tal human­i­ties com­mu­nity in the pre­vi­ous quarter.”

    digital-​​humanities jour­nal open-​​access pub­lish­ing
  • [1203.4881] Com­pu­ta­tional Com­plex­ity Analy­sis of Multi-​​Objective Genetic Programming

    Some days I just want to take genetic pro­gram­ming away from the com­puter sci­en­tists. Then I real­ize I ought to just let them keep the use­less, rit­u­al­ized thing they imag­ine it is.

    facepalm multiobjective-​​optimization software-​​development-​​is-​​not-​​programming
  • - How We Will Read: Clay Shirky

    “That is one of the poten­tial shifts in social read­ing: Can I cre­ate value for other peo­ple by say­ing that I found this pas­sage by Bruno LaTour strik­ing — even if I never look at it again? That’s an amaz­ing act of what I called “frozen shar­ing” in my last book. Being gen­er­ous about things when you are offer­ing it out to the pub­lic, with­out it being either in a spe­cific time frame or for a spe­cific target.”

    pub­lish­ing read­ing social-​​capital project be-​​useful-​​to-​​one-​​another

Items of some interest:

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

Items of some interest:

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