266 Matching Annotations
  1. Last 7 days
    1. Hsieh, C.-L., Goldsmith, J. A., Schaub, J. M., DiVenere, A. M., Kuo, H.-C., Javanmardi, K., Le, K. C., Wrapp, D., Lee, A. G., Liu, Y., Chou, C.-W., Byrne, P. O., Hjorth, C. K., Johnson, N. V., Ludes-Meyers, J., Nguyen, A. W., Park, J., Wang, N., Amengor, D., … McLellan, J. S. (2020). Structure-based design of prefusion-stabilized SARS-CoV-2 spikes. Science. https://doi.org/10.1126/science.abd0826

    1. Equivalent to role="region". Content that needs extra context from its parent sectioning element to make sense. This is a generic sectioning element that is used whenever it doesn’t make sense to use the other more semantic ones.
    2. Content that is self-contained in that it makes sense on its own when taken out of context. That could mean a widget, a blog post or even a comment within a blog post.
  2. Aug 2020
  3. Jul 2020
  4. Jun 2020
    1. Oh, and to be absolutely clear, every example that I know of a company trying to adopt the "pioneer - settler - town planner" structure without first getting their principles in a decent state has ... failed. Principles first. Org structure later. Culture, later still.

      Thinking on this with respect to Colin Woodard's book American Nations, makes me realize why America is in such horrible shape today.

    1. normalizing our dabatase will help us. What means normalize? Well, it simply means to separate our information as much as we can

      directly contradicts firebase's official advice: denormalize the structure by duplicating some of the data: https://youtu.be/lW7DWV2jST0?t=378

  5. May 2020
    1. We believe everyone deserves to report to exactly one person that knows and understands what you do day to day. The benefit of having a technically competent manager is easily the largest positive influence on a typical worker’s level of job satisfaction. We have a simple functional hierarchy, everyone has one manager that is experienced in their subject matter.
    2. We don't want a matrix organization where you work with a lead day to day but formally report to someone else.
  6. Apr 2020
    1. Par inhumain, nous entendons deux choses : premièrement, ce qui, par définition d’ordre épistémologique, n’est pas humain, soit du préfixe privatif in-, signifiant « sans » ;

      il manque une transition ici : je placerai ce paragraphe, un paragraphe avant, pour suivre l'idée d'un épuisement de l'anthropocentrisme et introduire Lautréamont

  7. Mar 2020
  8. Jan 2020
    1. Structuring instructional processes

      Examples below

      Amazing list!

    2. PROGRAMME STRUCTUREThe second set of variables that determine transactional distance are the elements in the course design, or the ways inwhich the teaching programme is structured so that it can be delivered through the various communications media.Programmes are structured in different ways to take into account the need to produce, copy, deliver, and controlthese mediated messages. Structure expresses the rigidity or flexibility of the programme's educational objectives,teaching strategies, and evaluation methods.
  9. Dec 2019
    1. I lost my mother's watch. And look! my last, or next-to-last, of three loved houses went.

      There is a subtle change from the third to the fourth stanza, a perfect split in keeping with the poem’s rigid structure. Almost imperceptibly, the speaker switches from addressing the reader to drawing on her own experience. It is here that Bishop begins to undermine her meticulous structural details and carefully impassive tone. “I lost my mother’s watch”, she states, an admission that seems to come from nowhere. However, the casual tone is disappearing; the inexplicable mention of this personal aspect of the speaker’s life has upped the emotional stakes. As the stanza continues, it becomes clear that this is a further attempt to demonstrate the universality of loss. The picture becomes bigger and the distance larger. The exclamation: “And look!” betrays yet more emotion, despite it’s apparent offhand tone. Now Bishop tells us to look at our losses on a bigger scale: the houses we lived in – not so disastrous except for the use of the word “loved” here. Indeed, these were just places we lived in, but we nonetheless also loved in them.

    2. —Even losing you (the joking voice, a gesture I love) I shan't have lied. It's evident the art of losing's not too hard to master though it may look like (Write it!) like disaster.

      he fifth stanza leads us to a brief look at the structure of the poem. The villanelle allows for a break in its pattern of tercets and tight rhyme, giving away to one quatrain with a repeated rhyme. Just as the structure cracks, as does the poetic voice. The final stanza opens with a dash, which could perhaps be seen as an attempt at a casual tone but in fact serves to slow the poem down here, allowing for yet more emotion to permeate the final words. The reader is forced to consider this “you”, and we see how the poem has taken a journey: starting with the little objects, going through thought and memory, to houses, places and continents forming one huge picture until at the end, zooming in on and pinpointing this “you”. A “you” with, as we infer from the parentheses, a personality, a memorable tone of voice and gestures. A person lost; an irreplaceable entity, in fact.

    1. Sec.??,

      Fix reference.

    2. Pass one of two single leptonb-performance triggers that require:444–An online reconstructed medium muon withpT> 26 GeV445–Or an online reconstructed electron with withpT> 26 GeV446•Contain an offline medium muon withpT> 30 GeV and no jet within∆R of 0.4.447•Contain an offline medium electron withpT> 30 GeV448•Contain at least two offline b-tagged jets defined as:449–Offline R=0.4 anti-kT jets.450–pT> 35 GeV and |η|< 2.5.451–Offline b-tagged at the 85% operating point.452–The offline jet must be matched to an online jet.

      In the bullet list, use semi-colons for the end of list items. Use periods only when completing a thought or sequence.

    3. Events are required are pass to the following selection

      "Events must pass the following selection:"

    4. The set of triggers with an electron or muon in it441withb-taggingalgorithm run on all online jets is called ‘b-performance trigger’ specifically used to study442b-jet trigger performance.

      This sentence is awkward. I recommend: "The set of trigger including an electron or muon along with the b-tagging algorithm is referred to as the 'b-performance trigger'. It is dedicated to the study of flavour-tagging performance in the HLT."

    5. electron plus a neutrino,

      "... electron and neutrino..." (and similarly in a few words, "... muon and neutrino ..."

    6. an internal documentation of flavour tagging note, [25],

      "... in an internal flavour tagging note (Ref.~\cite{})..."

    7. e,

      No comma - dependent clause follows.

    8. Event selection, data, and simulated samples are common to both methods.

      Avoid 1-sentence paragraphs. Combine this with the following text.

    9. representative combination

      "... representative combinations..."

    10. bTrig

      This notation ,with a capital "T" in the subscript, is clumsy. I recommend we just simplify this to $\epsilon_{\mathrm{trig}}$, since we are only talking about b-jet triggers here.

    11. spot

      "identify" ("spot" is too casual)

    12. A handy

      Too colloquial. "A guide to decoding..."

    13. denoted∗

      "... marked by a * ..." (this appears in a number of places - correct them all)

    14. denoted∗

      "... marked by a * ..."

    15. Tables??

      Fix broken reference.

    16. ≈40%

      Use $\sim$, not $\approx$, in this context.

    17. 70% 77%

      Missing comma in list.

    18. Further details of the performance and usage of the muon-jet chains inPbPbcollisions are detailed in365Appendix H and will be summarised here for the paper once we better understand from the HI experts366what plots they would like to have made public, how much detail they need us to cover here etc

      Do we have more information about what is desired for this part of the paper? It feels highly unfinished.

    19. HI r

      Was "HI" defined?

    20. to face the very

      "... were required to be robust against the high-multiplicity environment..."

    21. es, th

      "... small angles; this is the so-called..."

    22. so called


    23. golden probe

      This is a colloquial phrase. Define it, or use a more general term.

    24. muon jet

      Hyphenate "muon-jet"

    25. A

      Has "A" been defined? At the very least, cite the paper for this analysis here.

    26. bbb, a

      Missing parenthesis after"bbbb".

    27. s, a

      No comma - dependent clause.

    28. 1TeVand

      Space needed between "TeV" and "and"

    29. and

      Remove comma - a dependent clause follows.

    30. jets, and c

      No comma before "and"

    31. GSC

      ".. or GSC"

    32. Run 1 style

      "Run 1-style" -> hyphenate, since it's used as an adjective.

    33. he exact d

      "The exact" implies you have already discussed a target topology. Perhaps "An exact..." is better.

    34. , plus

      Replace ", plus" with " and ".

    35. anti-Kt


    36. volume that the tracking must be run o

      "... the volume on which tracking must be run ... "

    37. GeVwere

      Put a space after "GeV".

    38. Mean: 0.000098 +/- 0.000074RMS: 0.006637 +/- 0.000053Non-SplitMean: -0.000016 +/- 0.000090RMS: 0.007982 +/- 0.00006

      These font margins should be improved, so the text doesn't run into the right bounding box. This goes for all of these plots in Figure 4.

    39. and which follows closely the offline primary vertex finding strategy

      Make this phrase parenthetical

    40. weighted by found relative

      This phrase, "weighted by found relative," makes no sense. Please rephrase.

    41. inηandφ.

      "... in both eta and phi."

    42. downside of wasting

      How about "...inefficient use of CPU resources..." instead?

    43. selected

      "selected" or "produced"? Not sure what the intent is here for the underlying verb. ROIs would normally result (be produced) from the L1 trigger, then be used in the HLT. So is "produced" the right word?

    44. processed

      "Process" appears twice, once in the first two sentences. It has an over-use feel to it here. Replace one instance of "process" with an appropriate synonym.

    45. < μ >

      This looks like $<\mu>$. Instead, use $\langle \mu \rangle$. I recommend defining this in LATEX to simplify:

      \def\avgmu{\ensuremath{\langle \mu \rangle}}

    46. secondary

      Place secondary in quotes.

    1. He then showed you how he could make a few strokes on the keyset to designate the type of link he wanted established, and pick the two symbol structures that were to be linked by means of the light pen. He said that most links possessed a direction, i.e., they were like an arrow pointing from one substructure to another, so that in setting up a link he must specify the two substructures in a given order.
    2. "Most of the structuring forms I'll show you stem from the simple capability of being able to establish arbitrary linkages between different substructures, and of directing the computer subsequently to display a set of linked substructures with any relative positioning we might designate among the different substructures. You can designate as many different kinds of links as you wish, so that you can specify different display or manipulative treatment for the different types."
    3. "You usually think of an argument as a serial sequence of steps of reason, beginning with known facts, assumptions, etc., and progressing toward a conclusion. Well, we do have to think through these steps serially, and we usually do list the steps serially when we write them out because that is pretty much the way our papers and books have to present them—they are pretty limiting in the symbol structuring they enable us to use. Have you even seen a 'scrambled-text' programmed instruction book? That is an interesting example of a deviation from straight serial presentation of steps.3b6b "Conceptually speaking, however, an argument is not a serial affair. It is sequential, I grant you, because some statements have to follow others, but this doesn't imply that its nature is necessarily serial. We usually string Statement B after Statement A, with Statements C, D, E, F, and so on following in that order—this is a serial structuring of our symbols. Perhaps each statement logically followed from all those which preceded it on the serial list, and if so, then the conceptual structuring would also be serial in nature, and it would be nicely matched for us by the symbol structuring.3b6c "But a more typical case might find A to be an independent statement, B dependent upon A, C and D independent, E depending upon D and B, E dependent upon C, and F dependent upon A, D, and E. See, sequential but not serial? A conceptual network but not a conceptual chain. The old paper and pencil methods of manipulating symbols just weren't very adaptable to making and using symbol structures to match the ways we make and use conceptual structures. With the new symbol-manipulating methods here, we have terrific flexibility for matching the two, and boy, it really pays off in the way you can tie into your work.3b6d This makes you recall dimly the generalizations you had heard previously about process structuring limiting symbol structuring, symbol structuring limiting concept structuring, and concept structuring limiting mental structuring.
    4. As we are currently using it, the term includes the organization, study, modification, and execution of processes and process structures. Whereas concept structuring and symbol structuring together represent the language component of our augmentation means, process structuring represents the methodology component (plus a little more, actually). There has been enough previous discussion of process structures that we need not describe the notion here, beyond perhaps an example or two. The individual processes (or actions) of my hands and fingers have to be cooperatively organized if the typewriter is to do my bidding. My successive actions throughout my working day are meant to cooperate toward a certain over-all professional goal.
  10. Nov 2019
    1. Critique

      pourquoi la critique est-elle une sous-partie de la partie contenu ? Il y a ici un problème de niveau entre les titres stylo.

    2. C’est ce détail qui m’a le plus fait apprécié l’article de Fabrice Maroux.

      Il manque une conclusion ici qui permettrait de faire le lien entre la synthèse et la critique

    1. Article de Joanie Grenier (2017), extrait de la revue « Mémoires du livre ».

      Il aurait été attendu une présentation de l'auteure.

    1. Mais il a une méfiance des auteurs de la transparence du système de contrôle qui doit donc s’adapter. L’auteur questionne « les instances de validation » et la valorisation du contrôle par les « pairs » avec la contre-expertise. Il y a donc trois tendances : le libre accès à la science, la recherche contrôlée par les intermédiaires et les revues en libre accès qui sont créés par les chercheurs.

      le passage d'une idée à l'autre est ici un peu abrupt, manque de fluidité

    1. Ce texte remet en question la posture de l’écriture électronique, dans un contexte où le texte écrit est né en ligne à l’instar d’avoir été numérisé. L’auteur questionne aussi sur la valeur de l’écriture en ligne quand elle est placée en comparaison avec une écriture papier dite plus “conventionnelle”.

      Il aurait été nécessaire en introduction de présenter davantage l'article (édition et date de publication) ainsi que l'auteure (institution, champs de recherche)

    1. Désintermédiation ou « hyperintermédiation » ?

      cette partie critique, quoique très documentée, quitte un peu trop l'ouvrage et propose une enquête, qui demeure intéressante, du cas d'étude : nous perdons un peu de vue la structure et la logique d'un compte rendu critique

    2. Dans le cas présent, l’intermédiaire en question est le distributeur De Marque, entreprise fondée en 2009 et rapidement devenue le leader du livre numérique au Québec, grâce notamment à une entente avec l’Association nationale des éditeurs de livres (ANEL) et à la mise sur pied de l’Entrepôt numérique, espace virtuel abritant les contenus numériques qui sont ensuite offerts à la vente.

      la présentation du cas d'étude aurait pu être présente plus tôt

    3. dont certains ne sont plus fonctionnels - qu’il est possible de consulter directement d’un clic

      ce passage appartient davantage à la critique (absence de permaliens)

    4. Compte rendu critique de « De l’éditeur au lecteur : De Marque et la distribution du livre numérique ». Un texte de Joanie Grenier de l’Université de Sherbrooke, diffusé le 2 mai 2017 dans la revue Mémoires du livre, consulté sur Érudit, la plateforme universitaire québécoise de diffusion numérique.

      Ce passage aurait pu se trouver en résumé

    1. Un aspect rassurant L’aspect pédagogique du texte va ici de pair avec une certaine bienveillance vis-à-vis des lecteurs. En effet, à la conclusion, les auteurs mentionnent le côté « flottant » qu’a l’environnement numérique qui évolue de façon assez rapide. Apporter une vue globale sur le sujet de l’auctorialité permet ainsi de mieux le comprendre et de l’appréhender tout en brisant quelques idées présupposées.

      cette partie aurait pu être intégrée dans la partie précédente pour établir un point plus important.

    2. Méthodologie

      très bonne description synthètique ici

    1. Pour finir l’objectif est également de faire comprendre que l’édition numérique n’est que le produit d’une évolution culturelle et que sans le développement des téléphonnes, des ordinateurs, du web et d’Internet, l’édition numérique et électronique ne serait pas ce qu’elle est. Ce texte permet de mieux comprendre et mieux réaliser certains faits et la corrélation entre différents secteurs et évolutions au sein de la société. Un texte est toujours le reflet d’une époque.

      Ceci est votre conclusion et elle est très bonne

    2. L’objectif est également d’expliquer en quoi constitue cette nouvelle notion d’édition électronique, en expliquant point par point les trois principales dimensions de ce processus. Il est important d’expliquer cette notion car en compraraison aux méthodes d’éditions précédentes non numériques, elle vient se situer à la fois en continuité ainsi qu’en rupture de celles-ci. L’objectif de cet article est aussi de clarifier l’opinion publique et l’inquiétude que cette évolution a suscité quant aux différents rôles qui se voyaient émancipés dans ce processus numérique, Marin Daco et Pierre Mounier l’expliquent dans leur article en disant : […] on en effet pu laisser penser à une évolution rendant obsolète toute position intermédiaire entre le producteur (auteur) et le consommateur (lecteur). […] Plutôt qu’à une disparition de la position éditoriale, c’est à sa transformation qu’il faut penser. L’objectif est donc d’informer de la manière la plus objective possible et complète, la société, les lecteurs, les auteurs et les éditeurs quant à ce que représente l’édition électronique et ses nombreux et importants enjeux. L’objectif est également d’instruire la société quant à la révolution majeure de la notion de l’accessibilité et de la diffusion du savoir qu’occasionne l’édition électronique diffusée par le web sur Internet. C’est un rapport tout au autre au savoir qu’elle propose, ainsi qu’elle offre également cette notion de collaboration qui n’était guère possible avant.

      Cette partie aurait été placée dans la partie synthèse pour bien présenter les enjeux de leurs recherches.

    3. Objectifs et conclusion

      il est inhabituel de présenter les objectifs de l'ouvrage en conclusion....

    4. L’article Édition électronique qui nous intéresse ici, est comprit entre les pages 47 jusqu’à la page 55 de l’ouvrage Communication

      appartient davantage à l'introduction du document

  11. stylo.ecrituresnumeriques.ca stylo.ecrituresnumeriques.ca
    1. L’auteur démontre que le numérique s’inscrit dans un héritage de pratiques qu’il ne fait que réunir.

      bon point qui aurait mérité d'être en partie de synthèse

    2. Les deux premières parties tournent autour de la définition de la littérature électronique et du numérique. La troisième partie explique brièvement que le numérique se place dans une continuité de pratiques et caractéristiques qui existent depuis longtemps. Finalement, la quatrième partie propose pour chaque aspect culturel du numérique identifié par Vitali-Rosati, un aspect concordant dans l’œuvre de Plamondon (qui est un livre physique bien réel).

      Il aurait été préférable de procéder dans cette partie à une analyse chronologique : les allers-retours entre les différentes parties peuvent étourdir

    3. La quatrième partie et la première partie sont beaucoup plus longues que la deuxième et la troisième, qui elles ne prennent que quelques paragraphes. La deuxième et la troisième partie auraient gagné à être jumelées.

      Nous sommes déjà ici dans la critique

    4. Le texte souffre malheureusement de quelque inepties orthographiques et syntaxiques1. L

      Ce passage appartient à la critique

    1. Le texte expose une certaine tension entre la culture numérique avec l’idéalisme du partage et de la libre circulation du savoir et la culture imprimée qui est centrée sur l’objet livre (le livre papier). L’avènement du livre numérique affecte tous les secteurs gravitant de près ou de loin autour du livre. Il induit de prime abord une modification des habitudes de lecture qui suscite de l’enthousiasme chez certains et de la méfiance chez d’autres. La lecture immersive et linéaire (celle du livre relié) est mise à mal au profit de modes de lecture rapide et fragmentaire qui, semble-t-il, érige en modèle cognitif le déficit d’attention caractéristique des enfants difficiles. On parle de baisse du temps consacré à la lecture et on perçoit une menace à l’acte de lecture lui-même. Le monde éditorial est ébranlé par la dématérialisation des supports de diffusion et la pression vers la gratuité d’accès sur les biens informationnels. Il existe une certaine inquiétude face à la perspective de voir Google et Amazon, devenir les grands éditeurs de demain.

      Cette partie appartient davantage à la partie précédente.

    2. Propos du document

      Cette partie aurait mérité d'être davantage structurée avec un point sur les trois axes principaux de l'article, leurs limites, leurs remises en question, les phénomènes qui y sont liés.

  12. Oct 2019
  13. Sep 2019
    1. New York-based startup Tastemakers has raised a $1 million seed-round — led by Precursor Ventures — for its business that connects Africa adventures to global consumers.

      The first sentence states the main idea, so basically what the rest of the reading is going to be about

  14. Aug 2019
    1. Because of this distinction, outcome and impact evaluations are going to look a little different.

      This sentence might fit better, I think, if inserted earlier in the paragraph. To improve the general flow and logical presentation of important points, consider revising so that this paragraph looks something like:

      "When we talk about research or program evaluation, a lot of us tend to use the terms 'outcome' and 'impact' interchangeably. The truth, however, is that these are two distinct terms, and it's important to understand the differences between them. Because of these distinctions, outcome and impact evaluations are going to look a little different."

    2. In the end, it’s important to remember that these are guidelines, and you will no doubt encounter program evaluation projects that cross the lines of research, and vice versa. Understanding how the two differ will help you decide how to move forward when you encounter the need to assess the effect of a program in practice.

      Not sure if this was intentional or if this is an error, but I did want to mention: This paragraph ("In the end, it's important to remember that these are guidelines...") is a verbatim repetition of the concluding paragraph in the Execution subsection (right-arrow dropdown) above.

    1. Explain defining features of content analysis as a strategy for qualitative data analysis and when it is most effectively used

      These two learning objectives are perhaps best presented separately, since they're both important and both represent major goals for readers. Also, the way the sentence currently reads, it's not technically clear what is meant by "when it is most effectively used." Grammatically, "it" could refer to qualitative data analysis OR content analysis, depending on how the reader interprets this sentence. Consider revising to something like:

      "Learners will be able to:

      Explain defining features of content analysis as a strategy for analyzing qualitative data.

      Determine when content analysis can be most effectively used."

    2. A few exemplars of studies employing Thematic Analysis:

      I appreciate the author's inclusion of exemplars here. Is there a way to thread some of these exemplars into the main bodies of text in the Thematic Analysis sections? Some readers may tend to skip over these end-of-section details, in my experience.

    1. We are usually relatively unfamiliar with our participants, at least on a personal level. This can make sitting down for an interview where we might be asking some deep questions a bit awkward and uncomfortable, at least at first. Because of this, we want to craft our questions in such a way that they are not off-putting, inadvertently accusatory or judgmental, or culturally insensitive.  To accomplish this we want to make sure we phrase questions in a neutral tone (e.g. “Tell me what that was like”, as opposed to, “That sounds horrible, what was that like”). To accomplish this we can shift perspectives and think about what it would be like for us to be asked these questions (especially by a stranger), and we can pilot test our questions to see how they ‘feel’ to others. Also, if we are conducting interviews on topics that may be particularly hard for people to talk about, we likely will want to start out with some questions that are easier to address prior to getting into the heavier topics. Make them relatable Unlike surveys, where researchers may not be able to explain the meaning of question, when conducting interviews, we are present to help further explain questions if there is some confusion. However, ideally our questions are as clear as possible from the beginning. This means that we avoid jargon or technical terms, we anticipate areas that might be hard to explain and try to provide some examples or a metaphor that might help to get the point across, and we do our homework to relay our questions in a cultural context that is appropriate. Like the discussion above, pilot testing our questions can be very helpful for ensuring the relatability of our questions, especially with community representatives. What sounds good in our heads as a question, might make little sense to our intended audience. Make them individually distinct, but collectively comprehensive Just like when we are developing survey questions, you don’t want to ask more than one question at the same time. This is confusing and hard to respond to for the participant, so make sure you are only asking about one idea in each question.  However, when you are thinking about your list of questions, or your whole interview guide collectively, ensure that you have comprehensively included all the ideas related to your topic. It’s extremely disheartening as a qualitative researcher that has concluded their interviews and realized there was a really important area that you failed to include in your guide. To avoid this, make sure to know the literature in your area well and talk to other people who study this area to get there perspective on what topics need to be included.

      This table is a great review resource for student readers. However, perhaps the text can be formatted differently or broken up to facilitate quick review. Can we organize important points as bullet points rather than complete sentences? This may make for easier reader review.

    1. These two categories exist as opposite extremes on a continuum.

      I would suggest that we begin a new paragraph with the sentence "These two categories..."

      We may also want to very briefly reiterate the relevant topics. For example, the first sentence of the new paragraph might read something like "Idiographic and nomothetic research represent two different research categories existing at opposite extremes on a continuum."

    2. assure there is fair distribution of risks and benefits related to our research, be conscientious in our recruitment efforts to support equitable representation, and that we ensure special protections to vulnerable groups involved in research activities. As you plan your qualitative research study sampling plan, make sure to consider who is invited and able to participate and who is not.

      While this is certainly useful information, I think the sentence/paragraph structure may be a bit too dense to allow for adequate reader concept absorption. Is there a way to break down these 3 concepts into a more reader-friendly bulleted or numbered list? For example, instead of the current format/structure, we might present these topics in the following way:

      "Within this context, we need to:

      (1) Assure there is a fair distribution of risks and benefits related to our research;

      (2) Be conscientious in our recruitment efforts to support equitable representation; and

      (3) Ensure special protections for vulnerable groups involved in research activities."

      Then, perhaps we could begin a new paragraph with the sentence "As you plan your qualitative research study sampling method, make sure to consider..."

      I think the above might be helpful for readers who are better able to absorb content in smaller, more visually bite-sized chunks. Anyway, just a thought!

    3. across our participants.

      We might want to tweak the wording a bit here. Reading the sentence "...that reflects the ideas across our participants, I stumbled a bit when I ran into the word "across." Perhaps revise to something like "...reflects the ideas common among our participants" ?

  15. Jun 2019
    1. Federal Reserve System

      The Federal Reserve System is the central bank of the United States. It was founded by Congress in 1913 to provide the nation with a safer, more flexible, and more stable monetary and financial system.

  16. Apr 2019
    1. Il n'en reste pas moins que, du point de vue des relations plus que des représenta-tions, du dynamisme intellectuel.plus que des syntagmes, la ratiocina-tion est marquée par une engrammation dans l'imaginaire.

      Il est étonnant, et cela complexifie la lisibilité du propos, que autant de nature de points de vue entrent en compte dans une pensée de l'image en une seule phrase .... engrammation = biologique; ratiocination = rhétorique/littéraire; intellectuel etc. à moins que certains ne soient employés pour figurer une image ...


      Cette partie qui suit est étonnamment plus référencée que la première.

    3. mais au lieu de les affecter à un sens littéral, à un message d'ex-pression ou de communication, elle les inscrit dans de nouvelles combi-naisons linguistiques qui rendent active une pluralité de niveaux de signi-fications, entre lesquels s'instaure une coexistence, par libre jeu, de sens figurés.

      Toute la pensée de l'image chez l'auteur semble passer par une science de la parole et du mot.

    4. les images médiatement riches de pensée, parce qu'elles nécessitent une démarche interprétative pour livrer leur profon-deur de pensée

      catégorie 2

    5. les images immédiatement donatrices de savoirs, parce qu'elles laissent l'information effieurer sans délai à la surface des figures (formes spatiales et images verbales)

      catégorie 1

  17. Feb 2019
    1. Impact of Fully Connected Layers on Performance of Convolutional Neural Networks for Image Classification

      作者总结说:1)CNN 层越少,FC 层里的node 就要越多才行。相反 CNN 越深,FC node 少就够了;2)浅的 CNN 除了需要更多 FC node 外,数据集 class 类目数越多,FC 层应该越多越好,反之亦然;3)对于单个 class 内样本越多的数据集,网络越深越好,但若 class 类目数很多,浅的网络表现会更好。

  18. Nov 2018
    1. Deep convolutional Gaussian processes

      似乎很有趣,可惜我统计没学好,没掌握到背后的本质,需要回炉重造了[委屈]~ 文章居然用的是今年年初发表的 UMAP 进行降维可视化,而不是 t-sne,这很新颖嘛!

    2. Deep processing of structured data

      此文提出“set aggregation network”(SAN)子网络,但说白了是将卷积完的各特征图都做非线性化操作后再求和为一个图,so就有所谓pooling和flatten是其特例了。后文的实验还远算不上效果显著,甚至在我看来还没做完[挖鼻]~此文就随便看看得了~我这还没吐槽: 图像文本等不应该是“非结构化数据”嘛?[哼]

    3. On the loss landscape of a class of deep neural networks with no bad local valleys

      文章声称的全局最小训练,事实上主要基于一个比较特殊的人工神经网络的结构,用了各种连接到 output 的 skip connection,还有几个额外的assumptions 作为理论保证。

    4. How deep is deep enough? - Optimizing deep neural network architecture

      精心构造了一个度量:generalized discrimination value(GDV),实现了对网络各层中对不同输入类别的量化评价。这个评价定义还关于输入数据有平移不变性和缩放不变性 ,同时,并不依赖特征图的数目,也不依赖于每层中神经元的数目和排序。