A5039257513- Spectroscopy and Chemometric Analyses
- Building Energy and Comfort Optimization
- Remote-Sensing Image Classification
- Hygrothermal properties of building materials
- Wood Treatment and Properties
- Urban Heat Island Mitigation
- Spectroscopy Techniques in Biomedical and Chemical Research
- Control Systems and Identification
- Indoor and Outdoor Localization Technologies
- Sensory Analysis and Statistical Methods
- Tree Root and Stability Studies
- Video Surveillance and Tracking Methods
- Energy and Environment Impacts
- Remote Sensing in Agriculture
- Advanced Statistical Methods and Models
- Conservation Techniques and Studies
- Fault Detection and Control Systems
- Water Quality Monitoring and Analysis
- Solar Radiation and Photovoltaics
- Photovoltaic System Optimization Techniques
Malmö University
2021
Norwegian University of Life Sciences
2017-2020
Abstract A fast and memory‐efficient new method for performing genetic algorithm partial least squares (GA‐PLS) on spectroscopic data preprocessed in multiple different ways is presented. The method, which primarily intended datasets containing many observations, involves preprocessing a spectral dataset with several techniques concatenating the versions of horizontally into design matrix X both tall wide. large then condensed substantially smaller covariance T whose resulting size unrelated...
The hygrothermal surface conditions of a façade is important for the degradation material as well energy budget building. distribution short term variations temperature and moisture in often neglected studies that will typically treat whole equally. are especially porous building materials where water dependent biological physical processes main degrading factors. In this study diurnal cycle measured with conventional probes infra-red camera. addition, content wooden resistive measurement...
Abstract A novel formulation of the wide kernel algorithm for partial least squares regression (PLSR) is proposed. We show how elimination redundant calculations in traditional applications PLSR helps speeding up any choice cross‐validation strategy by utilizing precalculated lookup matrices. The proposed approach combined with some additional computational shortcuts resulting highly effective and numerically accurate results. advantages method are demonstrated comparisons to classical...
The purpose of this research is to develop a method for estimating the spatially and temporally resolved moisture content thermally modified Scots pine (Pinus sylvestris) using remote sensing. Hyperspectral time series imaging in NIR wavelength region (953–2516 nm) was used gather information about absorbance eight samples each minute as they dried during period approximately 20 h. After preprocessing collected spectral data identifying an appropriate selection, partial least squares...
It is recognised that flame retardant chemicals degrade and leach out of flame-protected wood claddings when exposed to natural weathering. However, the ability survey current state a treatment applied cladding, an arbitrary length time after initial application, limited today. In this study, hyperspectral imaging in near infrared short-wavelength region used quantify amount present on wooden surfaces. Several sets samples were treated with various concentrations chemical scanned push broom...
Feature selection is a challenging combinatorial optimization problem that tends to require large number of candidate feature subsets be evaluated before satisfying solution obtained. Because the computational cost associated with estimating regression coefficients for each subset, can an immensely time‐consuming process and often left inadequately explored. Here, we propose simple modification conventional sequence calculations involved when fitting same response data partial least squares...
Quantifying the number of occupants in an indoor space is useful for a wide variety applications. Attempts have been made at solving task using passive infrared (PIR) motion sensor data together with supervised learning methods. Collecting large labeled dataset containing both PIR and ground truth people count however time-consuming, often requiring one hour observation each gathered. In this paper, method proposed generating such synthetically. A simulator developed Unity game engine...