Superiority of blue (470nm) light in eliciting non-image forming brain responses during auditory working memory in humans : a fMRI study

Conference (2006)

Supernova 1987a in the Large Magellanic Cloud: an observational overview

Article for general public (1987)

Supernumerary Outer Hair Cells Arise External to the Last Row of Sensory Cells in the Organ of Corti

in Acta Oto-Laryngologica (2001), 121(2), 164-8

During the development of the mammalian inner ear, the number of hair cells produced is highly regulated and remains constant throughout life. The mechanism underlying this regulation is beginning to be ... [more ▼]

During the development of the mammalian inner ear, the number of hair cells produced is highly regulated and remains constant throughout life. The mechanism underlying this regulation is beginning to be understood although many aspects still remain obscure. When late embryonic or early postnatal rat organs of Corti were cultured, the production of supernumerary hair cells was observed. This overproduction of sensory cells could be modulated by the addition of several growth factors. In this study, we examined explants of rat organs of Corti that produced supernumerary hair cells. In the supernumerary hair cell region, up to two rows of inner hair cells and five rows of outer hair cells were observed. Morphological evaluation of these specimens revealed that less mature hair cells were located in the most external rows of these sensory cells. When a supernumerary hair cell was produced, a supporting cell (i.e. Deiters' cell) was also produced, strongly suggesting that the conversion of a Deiters' cell into a hair cell was not the mechanism that produced these extra hair cells. Based on these results, we propose that prosensory cells located at the external edge of the organ of Corti retain a capacity to form hair cells and that it is these prosensory cells that differentiate into supernumerary hair cells and Deiters' cells. [less ▲]

Superovulation in cows with FSH preparation containing different amounts of LH

in Roczniki Naukowe Zootechniki (1995), 22(2), 103-108

One of the main factors influencing the effectiveness of superovulation is the quality of gonadotropin used.Especially the proportion of FSH to Lh can play an important role.In the experiment ... [more ▼]

One of the main factors influencing the effectiveness of superovulation is the quality of gonadotropin used.Especially the proportion of FSH to Lh can play an important role.In the experiment superovulatory response of cows treated with FSH containing different amounts of LH was evaluated.Observations were carried out on 120 Polish Black-and-White Lowland cows aged from 5 to 8 years.On day 9, 10 and 11 after oestrus, 40 mg of FSH containing 10,20 or 40% of LH was injected in 8 decreasing doses at 12 h intervals.In the control group 3000 IU PMSG per cow was administrated.Forty-eight hours after the initial FSH or PMSG injection, 3 ml PGF 2-alpha analog was injected to donors to induce oestrus.An average of 5.6, 5.3, 3.9 and 1.7 transferable embryos was recovered after treatment of the cows with FSH +20% LH, FSH +40% LH, FSH +10% LH and PSMG, respectively. [less ▲]

Superovulation of goats with purified pFSH supplemented with defined amounts of pLH

in Theriogenology (1995), 43

The superovulatory response of goats treated with purified pFSH supplemented with 30, 40 or 50% pLH was compared. Sixty-four Boer goat does were synchronized by progestagen-containing ear implant ... [more ▼]

The superovulatory response of goats treated with purified pFSH supplemented with 30, 40 or 50% pLH was compared. Sixty-four Boer goat does were synchronized by progestagen-containing ear implant, randomly allotted to 3 groups and, beginning 2 d before implant removal, treated with purified pFSH supplemented with 30, 40 or 50% pLH. Each animal received 16 Armour Units of pFSH administered in 6 descending doses at 12-h intervals. Along with the last 2 injections, the does received 5 mg PGF(2alpha). Embryos were flushed either surgically or after slaughter on Day 5 or 6 after the last day of standing estrus. The percentage of animals responding to treatment was not different among groups treated with pFSH supplemented with 30, 40 or 50% pLH (76, 71 and 63%, respectively). The corresponding data for number of ovulations was 11.3 +/- 1.6, 16.3 +/- 1.8 and 16.4 +/- 2.6, for number of ova and embryos recovered 8.1 +/- 1.9, 12.0 +/- 1.5 and 13.5 +/- 2.9 and for number of transferable embryos 6.6 +/- 1.9, 9.1 +/- 1.5 and 7.1 +/- 2.1 (x +/- SEM). Results confirm the earlier finding of a good response of goats to pFSH preparations with a high FSH:LH ratio, and, although group differences were statistically nonsignificant (P > 0.05), they suggest that supplementation with approximately 40% pLH may be close to the optimum [less ▲]

A superposed epoch analysis of auroral evolution during substorms: Local time of onset region

in Journal of Geophysical Research. Space Physics (2010), 115

Previous workers have shown that the magnetic local time (MLT) of substorm onset depends on the prevailing east-west component of the interplanetary magnetic field (IMF). To investigate the influence of ... [more ▼]

Previous workers have shown that the magnetic local time (MLT) of substorm onset depends on the prevailing east-west component of the interplanetary magnetic field (IMF). To investigate the influence of the onset MLT on the subsequent auroral response we perform a superposed epoch analysis of the auroral evolution during approximately 2000 substorms using observations from the FUV instrument on the Imager for Magnetopause-to-Aurora Global Exploration (IMAGE) spacecraft. We subdivide the substorms by onset latitude and onset local time before determining average auroral images before and after substorm onset, for both electron and proton aurorae. We find that during the growth phase there is preexisting auroral emission in the MLT sector of the subsequent onset. After onset the auroral bulge expands eastward and westward, but remains centered on the onset sector. Approximately 30 min after onset, during the substorm recovery phase, the peaks in electron and proton auroral emission move into the postnoon and prenoon sectors, respectively, reflecting the “average” auroral precipitation patterns determined by previous studies. Superposed epoch analysis of the interplanetary magnetic field for the substorms under study suggests that the B[SUB]Y[/SUB] component of the IMF must be biased toward positive or negative values for up to a day prior to onset for the onset MLT to be influenced. [less ▲]

SUPERSPACE FORMULATION OF N=2 PSEUDOMECHANICS AND SUPERPOTENTIALS

in Journal of Mathematical Physics (1987), 28(2), 477-484

Superthermal particules in planetary atmospheres

in Non-stable processes in the Univers (1994)

Supervised learning for a Kraft recovery boiler: a data mining approach with Random Forests.

in Favrat, Daniel; Maréchal, François (Eds.) ECOS 2010 Volume IV (Power plants and Industrial processes) (2011, January 01)

A data mining methodology, the random forests, is applied to predict high pressure steam production from the recovery boiler of a Kraft pulping process. Starting from a large database of raw process data ... [more ▼]

A data mining methodology, the random forests, is applied to predict high pressure steam production from the recovery boiler of a Kraft pulping process. Starting from a large database of raw process data, the goal is to identify the input variables that explain the most significant output variations and to predict the high pressure steam flow. [less ▲]

Supervised Learning for Sequential and Uncertain Decision Making Problems - Application to Short-Term Electric Power Generation Scheduling

Doctoral thesis (2010)

Our work is driven by a class of practical problems of sequential decision making in the context of electric power generation under uncertainties. These problems are usually treated as receding horizon ... [more ▼]

Our work is driven by a class of practical problems of sequential decision making in the context of electric power generation under uncertainties. These problems are usually treated as receding horizon deterministic optimization problems, and/or as scenario-based stochastic programs. Stochastic programming allows to compute a first stage decision that is hedged against the possible futures and -- if a possibility of recourse exists -- this decision can then be particularized to possible future scenarios thanks to the information gathered until the recourse opportunity. Although many decomposition techniques exist, stochastic programming is currently not tractable in the context of day-ahead electric power generation and furthermore does not provide an explicit recourse strategy. The latter observation also makes this approach cumbersome when one wants to evaluate its value on independent scenarios. We propose a supervised learning methodology to learn an explicit recourse strategy for a given generation schedule, from optimal adjustments of the system under simulated perturbed conditions. This methodology may thus be complementary to a stochastic programming based approach. With respect to a receding horizon optimization, it has the advantages of transferring the heavy computation offline, while providing the ability to quickly infer decisions during online exploitation of the generation system. Furthermore the learned strategy can be validated offline on an independent set of scenarios. On a realistic instance of the intra-day electricity generation rescheduling problem, we explain how to generate disturbance scenarios, how to compute adjusted schedules, how to formulate the supervised learning problem to obtain a recourse strategy, how to restore feasibility of the predicted adjustments and how to evaluate the recourse strategy on independent scenarios. We analyze different settings, namely either to predict the detailed adjustment of all the generation units, or to predict more qualitative variables that allow to speed up the adjustment computation procedure by facilitating the ``classical'' optimization problem. Our approach is intrinsically scalable to large-scale generation management problems, and may in principle handle all kinds of uncertainties and practical constraints. Our results show the feasibility of the approach and are also promising in terms of economic efficiency of the resulting strategies. The solutions of the optimization problem of generation (re)scheduling must satisfy many constraints. However, a classical learning algorithm that is (by nature) unaware of the constraints the data is subject to may indeed successfully capture the sensitivity of the solution to the model parameters. This has nevertheless raised our attention on one particular aspect of the relation between machine learning algorithms and optimization algorithms. When we apply a supervised learning algorithm to search in a hypothesis space based on data that satisfies a known set of constraints, can we guarantee that the hypothesis that we select will make predictions that satisfy the constraints? Can we at least benefit from our knowledge of the constraints to eliminate some hypotheses while learning and thus hope that the selected hypothesis has a better generalization error? In the second part of this thesis, where we try to answer these questions, we propose a generic extension of tree-based ensemble methods that allows incorporating incomplete data but also prior knowledge about the problem. The framework is based on a convex optimization problem allowing to regularize a tree-based ensemble model by adjusting either (or both) the labels attached to the leaves of an ensemble of regression trees or the outputs of the observations of the training sample. It allows to incorporate weak additional information in the form of partial information about output labels (like in censored data or semi-supervised learning) or -- more generally -- to cope with observations of varying degree of precision, or strong priors in the form of structural knowledge about the sought model. In addition to enhancing the precision by exploiting information that cannot be used by classical supervised learning algorithms, the proposed approach may be used to produce models which naturally comply with feasibility constraints that must be satisfied in many practical decision making problems, especially in contexts where the output space is of high-dimension and/or structured by invariances, symmetries and other kinds of constraints. [less ▲]