Papers by Jeanette Hariharan
Biosystems Engineering, Sep 30, 2023
IntechOpen eBooks, May 17, 2023
This chapter introduces several feature extraction techniques (FETs) and machine learning algorit... more This chapter introduces several feature extraction techniques (FETs) and machine learning algorithms (MLA) that are useful for pattern recognition in hyperspectral data analysis (HDA). This chapter provides a handbook of the most popular FETs that have proven successful. Machine learning algorithms (MLA) for use with HDA are becoming prevalent in pattern recognition literature. Several of these algorithms are explained in detail to provide the user with insights into applying these for pattern recognition. Unsupervised learning applications are useful when the system is provided with the correct set of independent variables. Various forms of linear regression assay adequately solve hyperspectral pattern resolution for identifying phenotypes. K-means is an unsupervised learning algorithm that is used for systematically dividing a dataset into K number of pattern groups. Supervised and unsupervised neural networks (NNs) are used to discern patterns in hyperspectral data with features as inputs and in large datasets where little a priori knowledge is applied. Other supervised machine learning procedures derive valuable feature detectors and descriptors through support vector machine. Several methods using reduced sets for extracting patterns from hyperspectral data are shown by discretized numerical techniques and transformation processes. The accuracy of these methods and their usefulness is generally assessed.
Social Science Research Network, 2022
Hyperspectral Imaging - A Perspective on Recent Advances and Applications [Working Title]
This chapter introduces several feature extraction techniques (FETs) and machine learning algorit... more This chapter introduces several feature extraction techniques (FETs) and machine learning algorithms (MLA) that are useful for pattern recognition in hyperspectral data analysis (HDA). This chapter provides a handbook of the most popular FETs that have proven successful. Machine learning algorithms (MLA) for use with HDA are becoming prevalent in pattern recognition literature. Several of these algorithms are explained in detail to provide the user with insights into applying these for pattern recognition. Unsupervised learning applications are useful when the system is provided with the correct set of independent variables. Various forms of linear regression assay adequately solve hyperspectral pattern resolution for identifying phenotypes. K-means is an unsupervised learning algorithm that is used for systematically dividing a dataset into K number of pattern groups. Supervised and unsupervised neural networks (NNs) are used to discern patterns in hyperspectral data with features ...
Construction Research Congress 2022, 2022
2020 IEEE Green Technologies Conference(GreenTech), 2020
In the era of information explosion, cloud technology is the backbone of data storage. The role o... more In the era of information explosion, cloud technology is the backbone of data storage. The role of a continuous and reliable cloud system is crucial in managing the widespread emergence of “big data”. Electronic data devices are generating data at an unprecedented rate that needs to be stored and managed. Data Centers (DCs) provide a place that is primarily designed for data storage and processing purposes. The constant growth in the number of data centers created subsequent environmental impacts including carbon emissions, resource depletion, and energy and material wastes. Assessing data centers sustainability is now provided by a few third-party verification systems. The USGBC LEED as the most prominent green building assessment system in the US, added a category for data centers in the current version LEED version 4 (LEED v4). A comprehensive review of the LEED v4 scoring criteria reveals that it fails to properly address and assess the unique attributes of data centers. LEED focuses on improving sustainability measures mainly in “Energy and Atmosphere” but pays less attention to other categories. A major flaw in LEED v4 approach to data centers is that LEED assesses similar scoring criteria to data centers as other building types. This study concludes with discussing the potential areas of improvement in LEED v4 for assessing sustainability measures in data centers.
Estimating future costs of construction is an important component to the success of any contracti... more Estimating future costs of construction is an important component to the success of any contracting company. Traditionally a cost modifier has been utilized to offset cost escalations or volatility predictions. Construction estimators and contractors have also attempted to utilize a variety of prediction models. This paper establishes a basis for reliable forecasting and explores the possibility of developing prediction models using time series Neural Networks (NN) by utilizing historic data of three accepted macroeconomic composite indicators (MEI) and two accepted construction industry cost indices. The use of these macroeconomic indicators for NN-based models may be used to predict cost escalations for construction. Nonlinear autoregressive NN models are constructed through using the macroeconomic data and the construction cost data to determine if a reliable time-series predictive model could be established. The results of these models indicated that there is a high correlation between the macroeconomic escalations, independent factors, and the construction cost escalations, dependent factors, over time. Use and knowledge of these correlations could aid in the prediction of cost escalations during construction.
Remote Sensing, 2019
Laurel wilt (Lw) is a very destructive disease and poses a serious threat to the commercial produ... more Laurel wilt (Lw) is a very destructive disease and poses a serious threat to the commercial production of avocado in Florida, USA. External symptoms of Lw are similar to those that are caused by other diseases and disorders. A rapid technique to distinguish Lw infected avocado from healthy trees and trees with other abiotic stressors is presented in this paper. A novel method was developed to analyze data from hyperspectral data using finite difference approximation (FDA) and bivariate correlation (BC) to discriminate Lw, Nitrogen (N), and Iron (Fe) deficiencies from healthy avocado plants. Several combinatorial methods were used in preprocessing the data, such as standard normal transformation of data, smoothing of the data, and polynomial fit. The FDA technique was derived using a Taylor Polynomial finite difference approximation. This FDA accentuates inflection points in the spectrum. These, in turn, reveal variance in the data that can be used to identify spectral signature asso...
2020 10th Annual Computing and Communication Workshop and Conference (CCWC), 2020
The objective of this paper is a novel interpretation of the spectral and imaging data analysis p... more The objective of this paper is a novel interpretation of the spectral and imaging data analysis process which takes into account the measurement of the variance caused by disease infestationof a cell. Using multivariate analysis, the Karhounen-LoeveExpansion(KLE) of hyperspectral reflectance data, taken from healthy and diseased states of several plant species, is used to identify a basis setof functions which represent the distribution of reflected signal energy. By spectral decomposition, the eigenvalues are related to the KLEbasis set. The eigenvalues can be used to identify the KLE eigenvectorswhich comprise the highest variation in the data. These componentscan be interpreted as the weighted variables which carry with themmost of the information on the reflectance spectrum of the cell. Fromindications presented by this multivariate KLE analysis, a frequencyreconstruction is adapted to convert the eigenvector information to awave function. This reconstruction via KLE and frequen...
Finite Difference Analysis and Bivariate of Hyperspectral Data for Detecting Laurel Wilt Disease and Nutritional Deficiency in Avocado, 2019
Laurel wilt (Lw) is vectored by the red bay ambrosia beetle and can kill species of the Lauraceae... more Laurel wilt (Lw) is vectored by the red bay ambrosia beetle and can kill species of the Lauraceae family, including avocado trees, within a few months. It is a very destructive disease, which poses a serious threat to the commercial production of avocado in Florida, USA. External symptoms of Lw are similar to those that are caused by other diseases and disorders. Herein, a rapid technique to distinguish Lw infected avocado from healthy trees and trees with other abiotic stressors (e.g., N and Fe deficiencies), utilizing hyperspectral imaging, and signal processing techniques is presented. To determine pattern information in spectral data, finite difference approximation is often used to detect underlying gradients in pattern frequencies. This work presents a novel technique that analyzes spectral data from hyperspectral imagery using a second order difference formula, and statistical correlation to discriminate Lw, Nitrogen (N) and Iron (Fe) deficiencies from healthy avocado plants. Several combinatorial methods were used in preprocessing the data, such as standard normal transformation of data, and smoothing of the data by a polynomial fit. Then using a Taylor Polynomial derived difference approximation, second order spectra were uncovered. The finite difference method accentuates nonlinearities in the spectrum. These, in turn, reveal variance in the data that can be used to identify signature spectra associated with healthy and diseased states. By statistical correlation of these spectral patterns, an algorithm for distinguishing Lw avocado leaves from all other categories of healthy or mineral deficient avocado leaves is achieved with an overall accuracy of nearly 100%.
2020 IEEE Green Technologies Conference, 2020
In the era of information explosion, cloud technology is the backbone of data storage. The role o... more In the era of information explosion, cloud technology is the backbone of data storage. The role of a continuous and reliable cloud system is crucial in managing the widespread emergence of "big data". Electronic data devices are generating data at an unprecedented rate that needs to be stored and managed. Data Centers (DCs) provide a place that is primarily designed for data storage and processing purposes. The progressive demand for cloud storage requires larger and more efficient data centers. The constant growth in the number of data centers created subsequent environmental impacts including carbon emissions, resource depletion, and energy and material wastes. Assessing data centers sustainability is now provided by a few third-party verification systems including the US Green Building Council (USGBC) Leadership in Energy and Environmental Design (LEED). The USGBC LEED as the most prominent green building assessment system in the US, added a category for data centers in the current version LEED version 4 (LEED v4). A comprehensive review of the LEED v4 scoring criteria reveals that it fails to properly address and assess the unique attributes of data centers. LEED focuses on improving sustainability measures mainly in "Energy and Atmosphere" but pays less attention to other categories. A major flaw in LEED v4 approach to data centers is that LEED assesses similar scoring criteria to data centers as other building types. This study concludes with discussing the potential areas of improvement in LEED v4 for assessing sustainability measures in data centers.
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020
This study introduces a novel method for extracting reflectance signatures from plants for identi... more This study introduces a novel method for extracting reflectance signatures from plants for identification and disease detection. The Karhounen-Loeve Expansion (KLE) of spectral reflectance data, taken from healthy and diseased plants, is used to identify a basis set of functions which represent the distribution of the reflected signal energy. By spectral decomposition, the eigenvalues are related to the KLE basis set. The eigenvalues can be used to identify the KLE eigenvectors which comprise the highest variation in the data. These components can be interpreted as the weighted variables which carry with them most of the information on the reflectance spectrum of the plant. From indications presented by this multivariate KLE analysis, a frequency reconstruction is adapted to convert the eigenvector information to a wave function. This reconstruction via KLE and frequency transformation forms the signature identification process. These frequency spectra can be used as average signature reflectance patterns for plant identification, classification and biomarkers for diseases. The defining of these spectral identification biomarkers or signatures is purposeful since it could lead to less invasive techniques for classification and disease diagnostics. The techniques used to determine these reflectance spectra require a unique and rarely used frequency separation method.
The Basis for Development of a Foundational Biomarker Reflectance Signature Database System for Plant Cell Identification, Disease Detection, and Classification Purposes , 2020
The objective of this paper is a novel interpretation of the spectral and imaging data analysis p... more The objective of this paper is a novel interpretation of the spectral and imaging data analysis process which takes into account the measurement of the variance caused by disease infestation of a cell. Using multivariate analysis, the Karhounen-Loeve Expansion (KLE) of hyperspectral reflectance data, taken from healthy and diseased states of several plant species, is used to identify a basis set of functions which represent the distribution of reflected signal energy. By spectral decomposition, the eigenvalues are related to the KLE basis set. The eigenvalues can be used to identify the KLE eigenvectors which comprise the highest variation in the data. These components can be interpreted as the weighted variables which carry with them most of the information on the reflectance spectrum of the cell. From indications presented by this multivariate KLE analysis, a frequency reconstruction is adapted to convert the eigenvector information to a wave function. This reconstruction via KLE and frequency transformation forms the signature identification process for developing a database of healthy cell reflectance pattern features and variations produced by disease or other factors. These frequency spectra can be used as average signature reflectance patterns for cell identification, classification and biomarkers for diseases. The defining of these spectral identification biomarkers or signatures, is purposeful since it could lead to less invasive techniques for classification and disease diagnostics. The techniques used to determine these reflectance spectra require a unique and rarely used transformation method. These processes need further testing and verification through multiple refinements of this procedure.
Uploads
Papers by Jeanette Hariharan