La ridgeregression a été introduite pour résoudre des problèmes de multicolinéarité dans la régression multiple. Le calcul de stockage de l'eau du lac Okeechobee, Floride sert d'exemple pour présenter la technique et ses applications. Comparaison
de la performance de cette méthode par rapport à celle obtenue par la technique des moindres carrés. Les résultats de la ridgeregression montre que l'écart par rapport aux mesures de stockage n'est pas seulement affecté par les apports en eau (comme
AMS radiocarbon dating of Lake Michigan beach-ridge and dune development
C 14 dating ; Dune ; Lake ; Lake level ; Michigan ; Palaeo-environment ; Palaeoclimate ; Regression analysis ; United States of America
Strandplains of shore-parallel beach ridges bordering the Great Lakes are valuable for reconstructing histories of climate-related lake-level fluctuations. The resolution and imprecision of radiocarbon chronologies can be improved with AMS C 14
dates of roots and rhizomes of plant species associated with the formation and growth of the sand -dune caps of beach ridges. An AMS radiocarbon chronology of beach-ridge formation in northern Lake Michigan shows that information about variation
in the frequency of ridge formation is important for paleoclimatic interpretation.
The generalized ridge estimator is described under assumptions of normality, as a Bayesian estimator. The properties of this estimator and its derivatives (ridge estimator, shrunken estimator, Marquardt's estimator) which are biased estimators
of the regression coefficients vector, are studied together with the criterion of mean square error.
Chronosequence ; Holocene ; Lake ; Lake level ; Michigan ; Pedogenesis ; Podsolisation ; Regression analysis ; Soil properties ; United States of America
of the ridges, with an age range of 10 to 5200 years. Chronofunctions were developed using statistical regression techniques. Discussion of the results.
A sequence of approximately 75 strand plain beach ridges located along the northern shore of Lake Michigan, USA, was used to explore rates of soil development in the initial stages of podzolisation. Well drained pedons were sampled on 24
Les AA. présentent les premiers résultats d'une étude comportant l'application de différentes méthodes en vue de la prévision des étiages de l'Oise à Sempigny: méthodes de régression (Ridgeregression, Stepwise et Cross validation), méthodes
Lorsqu'on utilise des modèles précipitation-débit de type régressif, on peut être confronté à des problèmes de corrélation indésirable entre les variables explicatives, dus à la collinéarité. On peut les traiter par ridgeregression. Justification
Analyse spatiale ; Distribution ; Estimation ; Généralités sur la géographie ; Moindres carrés ; Méthode de Gram-Schmidt ; Point ; Ridgeregression ; Statistique ; Surface de tendance
. La ridgeregression ne semble pas être valable dans ce cas| la méthode des polynômes orthogonaux, à partir du procédé de Gram-Schmidt donne de meilleurs résultats que les moindres carrés.
Subsurface data from the Miranda coastal plain, Firth of Thames, confirm that shell ridges between Kaiana and Miranda are true cheniers. Surveys of a modern chenier indicate that coarse sediment and shell form sand-bars on the foreshore
and that these migrate landwards through swash action. The continuity of the shell layer, radiometrically dated to 3650 60 yrs BP to landward and 1260 50 yrs BP to seawards implies that there have not been oscillations of sea-level during this regressive coastal phase
while the shell ridges have developed. Mactra were also found in situ at the north and south of the coastal plain dated to 4510 70 and 200 60 yrs BP respectively.
Observations from the manned apparatus Sever-2 revealed ancient landforms due to marine erosion and subaerial processes in the Whale Ridge at the depth of 220 to 895 m. The landforms are thought to be formed in the Miocene during the planetary ocean
regression resulting from the glaciation in Antarctic. - (L'Ed.).
High-mountain valley floors evolution during regression of alpine glaciers in the Massif des Ecrins, France
The age of recent glacial, glacifluvial/fluvial forms is determined using lichenomeric dating in 2 Alpine valleys of the Massif des Ecrins. A lichen growth-curve is based on data from natural sites, such as boulder-fields, moraine ridges etc
Modeling stream-bank erosion in the Southern Blue Ridge Mountains
Bassin-versant ; Bilan sédimentaire ; Blue Ridge ; Cours d'eau ; Erosion des berges ; Etats-Unis ; Géomorphométrie ; Modèle ; Montagne ; North Carolina ; Plaine d'inondation ; Terrasse fluviatile ; Vitesse d'érosion
with different land covers (Coweeta Creek and Skeenah Creek watersheds in the Upper Little Tennessee River basin). Bank-derived sediment yield estimates were modeled in a Geographic Information System, using linear regression to relate floodplain widths
Twenty localities were sampled and the corresponding sedimentary sections are given. The areal distribution and thickness of the shell deposits are shown. These wedge-shaped ridges have been left as beach deposits by a regression of the sea. The C
Soil organic carbon (SOC) in a humid subtropical forest in Puerto Rico is higher at ridge locations compared to valleys, and therefore opposite to what is commonly observed in other forested hillslope catenas. To better understand the spatial
dynamics and soil chemistry. To investigate the influence of litter dynamics, the Century soil organic model was parameterized to simulate the effect of substituting valley species for ridge species. Soil chemical controls on C concentrations were
investigated with multiple linear regression models using iron, aluminum and clay variables.
DISTRIBUTION SPATIALE DU REVENU. UNE EXPERIENCE D'APPLICATION DE LA REGRESSION PAR PALIERS EN GEOGRAPHIE
ANALYSE DE L'ESPACE ; Géographie de l'Europe ; ITALIE ; LOMBARDIE ; METHODOLOGIE ; REGRESSION PAS A PAS ; REVENU
LA METHODE DE REGRESSION PAS A PAS (STEPWISE REGRESSION) EST APPLIQUEE A L'ETUDE DES REVENUS MONETAIRES. TROIS REGRESSIONS ONT ETE EFFECTUEES, D'ABORD AVEC TROIS GROUPES DIFFERENTS D'INDICATEURS DU REVENU, UNE SECONDE EN UTILISANT LES PREMIERES
VARIABLES RETENUES DANS LES 3 PRECEDENTES REGRESSIONS ET UNE TROISIEME, TOUJOURS AVEC CES DERNIERES, APRES AVOIR MODIFIE LES VALEURS DU Y, SUIVANT L'IMPORTANCE ET LE SIGNE DES RESIDUS.