Defending Grey Attacks by Exploiting Wavelet Analysis in Collaborative Filtering Recommender Systems

  • - Z
  • Cai Z
  • Esmaeilikelishomi A
N/ACitations
Citations of this article
5Readers
Mendeley users who have this article in their library.

Abstract

The identification of spatial and temporal patterns of soil properties and moisture structures is an important challenge in environmental and soil monitoring as well as for soil landscape model approaches. This work examines the use of hyperspectral remote sensing techniques for quantifying geophysical parameters from the hyperspectral reflectance of the vegeta-tion canopy. These can be used as proxies of the underlying soil and soil water conditions. Different spectral index derivatives, single band reflectance, and spectral indices from the airborne hyperspectral sensor AISA were quantified and tested in univariate and multivariate regression models for their correlation with geophysical measurements with electromagnetic induction (EMI) and gamma-ray spectrometry. The best univariate models for predicting elec-trical conductivity based on spectral information were based on the vertical dipole EM38DD V with an R 2 = 0.54 with the spectral index Normalized Pigments Reflectance Index (NPCI) as well as for the horizontal dipole EM38DD H with an R 2 = 0.65 with the spectral index NPCI. For predicting soil characteristics measured with gamma-ray spectrometry we received the best model results for gTh with an R 2 = 0.55 with the spectral index NPCI and gK with an R 2 = 0.44 with the spectral index Triangular Vegetation Index (TVI) and NPCI. The combination of variables including the geographical elevation was tested as the input for a multivariate regression analysis. For EMI and gamma-ray measurements, the " elevation " was found to be the most predictive variable and an integration of spectral indices into the elevation-based model led to only a slight improvement in the predictive power for EMI. An improvement could be made to explain the variance of gamma-ray measurement signals by combining elevation and spectral information. Abbreviations: CAI, cellulose absorption index; NPCI, normalized pigments reflectance index; PLS, partial least-squares; PSRI, plant senescence reflectance index; QP, quadrature phase; SWIR, short wave infrared; TVI, triangular vegetation index. Soil heterogeneity is a key challenge when modeling the flow, transport, and turnover processes in the soil-landscape context. Reliable forecasts require profound knowledge about the variability of soil parameters and the functional heterogeneity in terrestrial systems. Therefore, methods are required to measure and assess the distribution and pat-tern of soil properties. Quantifying and qualifying the spatial and temporal patterns of soil properties and mois-ture characteristics is still one of the central challenges in environmental monitoring. An adequate description of soil variability is an essential piece of information to put into ecological modeling, agriculture, and soil management (Bouma et al., 1999; Grayson and Blöschl, 2001; Lin et al., 2005, 2006; Schulz et al., 2006). Soil maps are the key to pro-viding information about soil distribution, soil structures, underlying processes on scales appropriate for the modeling or management of soils and for linking, monitoring data, and understanding landscape characteristics (Bouma, 2009; Heuvelink and Webster, 2001; Lin et al., 2005; McBratney et al., 2003; Scull et al., 2003; van Egmond et al., 2009). Soil characteristics and soil moisture patterns are important site conditions affecting bio-chemical–physical properties of plants and vegetation as a result of adaptation or plant stress. (Feilhauer and Schmidtlein, 2011; King et al., 2012; Schmidtlein and Sassin, 2004). The functional reactions of plants and vegetation are controlled and influenced by a com-bination of soil properties including characteristics such as texture, salinity, pH-level, chemical composition, soil moisture patterns, and temperature (Tromp-van Meerveld and McDonnell, 2009; Li et al., 2011; Schmidtlein et al., 2012). The presence of functional

Cite

CITATION STYLE

APA

-, Z., Cai, Z., & Esmaeilikelishomi, A. (2015). Defending Grey Attacks by Exploiting Wavelet Analysis in Collaborative Filtering Recommender Systems. International Journal of Advanced Research in Artificial Intelligence, 4(11). https://doi.org/10.14569/ijarai.2015.041103

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free