Principal Support Research Articles

Facile PEG-based isolation and classification of cancer extracellular vesicles and particles with label-free surface-enhanced Raman scattering and pattern recognition algorithm.

Extracellular vesicles and particles (EVPs), which contain the same surface proteins as their mother cells, are promising biomarkers for cancer liquid biopsy. However, most of the isolation methods of EVPs are time-consuming and complicated, and hence, sensitive detection and classification methods are required for EVPs. Here, we report a facile polyethylene glycol (PEG)-based method for isolating and classifying EVPs with label-free surface-enhanced Raman scattering (SERS) and pattern recognition algorithm. There are only three steps in the PEG-based isolation method, and it does not require ultracentrifugation, which makes it a low-cost and easy-to-use method. Three types of common male cancer cell lines, namely leukemia (THP-1), prostate cancer (DU-145), and colorectal cancer (COLO-205), and one healthy male blood sample, were utilized to isolate EVPs. To collect the SERS spectra of EVPs, a novel planar nanomaterial, namely amino molybdenum oxide (AMO) nanoflakes, was applied, with the enhancement factor being obtained as 3.2 × 102. Based on the principal component analysis and support vector machine (PCA-SVM) algorithm, cancer and normal EVPs were classified with 97.4% accuracy. However, among the cancer EVPs, the accuracy, precision, and sensitivity were found to be 90.0%, 90.9%, and 83.3% for THP-1; 86.7%, 80.0%, and 92.3% for DU-145; 96.7%, 83.3%, and 100% for COLO-205, respectively. Thus, this work will improve the isolation, detection, and classification of EVPs and promote the development of cancer liquid biopsies.

Identifying factors contributing to the development and sustainability of professional learning communities in Malaysian secondary schools

The aim of this research was to develop a Professional Learning Communities Model in Malaysian secondary schools. Exploratory Factor Analysis (EFA) was conducted to identify the underlying factors whereas Confirmatory Factor Analysis (CFA) was used to examine the measurement models. A total of 532 and 1062 respondents completed the survey with usable data for EFA and CFA, respectively. With normed χ² = 3.223, TLI = .950, CFI = .953 and RMSEA = .046, the results suggested that the model can be explained by two factors: Organisational Factor and Non-Organisation Factor. The Organisational Factor consists of four indicators: Shared Norms and Vision, Principal Commitment and Support, Structural Support, and Collegial Understanding and Trust whereas Non-Organisational Factor encompasses four indicators: Collaborative Learning, Reflective Dialogue, Collective Inquiry and External Support System.

Voice Feature Extraction for Gender and Emotion Recognition

Voice recognition plays a key function in spoken communication that facilitates identifying the emotions of a person that reflects within the voice. Gender classification through speech is a popular Human Computer Interaction (HCI) method on account that determining gender through computer is hard. This led to the development of a model for "Voice feature extraction for Emotion and Gender Recognition". The speech signal consists of semantic information, speaker information (gender, age, emotional state), accompanied by noise. Females and males have specific vocal traits because of their acoustical and perceptual variations along with a variety of emotions which bring their own specific perceptions. In order to explore this area, feature extraction requires pre-processing of data, which is necessary for increasing the accuracy. The proposed model follows steps such as data extraction, pre-processing using Voice Activity Detector(VAD), feature extraction using Mel-Frequency Cepstral Coefficient(MFCC), feature reduction by Principal Component Analysis(PCA) and Support Vector Machine (SVM) classifier. The proposed combination of techniques produced better results which can be useful in healthcare sector, virtual assistants, security purposes and other fields related to Human Machine Interaction domain.

Multiple microarray analyses identify key genes associated with the development of Non-Small Cell Lung Cancer from Chronic Obstructive Pulmonary Disease.

Introduction: Chronic obstructive pulmonary disease (COPD) is an independent risk factor of non-small cell lung cancer (NSCLC). This study aimed to analyze the key genes and potential molecular mechanisms that are involved in the development from COPD to NSCLC.Methods: Expression profiles of COPD and NSCLC in GSE106899, GSE12472, and GSE12428 were downloaded from the Gene Expression Omnibus (GEO) database, followed by identification of the differentially expressed genes (DEGs) between COPD and NSCLC. Based on the identified DEGs, functional pathway enrichment and lung carcinogenesis-related networks analyses were performed and further visualized with Cytoscape software. Then, principal component analysis (PCA), cluster analysis, and support vector machines (SVM) verified the ability of the top modular genes to distinguish COPD from NSCLC. Additionally, the corrections between these key genes and clinical staging of NSCLC were studied using the UALCAN and HPA websites. Finally, a prognostic risk model was constructed based on multivariate Cox regression analysis. Kaplan-Meier survival curves of the top modular genes on the training and verification sets were generated.Results: A total of 2350, 1914, and 1850 DEGs were obtained from GSE106899, GSE12472, and GSE12428 datasets, respectively. Following analysis of protein-protein interaction networks, the identified modular gene signatures containing H2AFX, MCM2, MCM3, MCM7, POLD1, and RPA1 were identified as markers for discrimination between COPD and NSCLC. The modular gene signatures were mainly enriched in the processes of DNA replication, cell cycle, mismatch repair, and others. Besides, the expression levels of these genes were significantly higher in NSCLC than in COPD, which was further verified by the immunohistochemistry. In addition, the high expression levels of H2AFX, MCM2, MCM7, and POLD1 correlate with poor prognosis of lung adenocarcinoma (LUAD). The Cox regression prognostic risk model showed the similar results and the predictive ability of this model is independent of other clinical variables.Conclusions: This study revealed several key modules that closely relate to NSCLC with underlying disease COPD, which provide a deeper understanding of the potential mechanisms underlying the malignant development from COPD to NSCLC. This study provides valuable prognostic factors in high-risk lung cancer patients with COPD.

An SVM-Based Nested Sliding Window Approach for Spectral–Spatial Classification of Hyperspectral Images

This paper proposes a Nested Sliding Window (NSW) method based on the correlation between pixel vectors, which can extract spatial information from the hyperspectral image (HSI) and reconstruct the original data. In the NSW method, the neighbourhood window constructed with the target pixel as the centre contains relevant pixels that are spatially adjacent to the target pixel. In the neighbourhood window, a nested sliding sub-window contains the target pixel and a part of the relevant pixels. The optimal sub-window position is determined according to the average value of the Pearson correlation coefficients of the target pixel and the relevant pixels, and the target pixel can be reconstructed by using the pixels and the corresponding correlation coefficients in the optimal sub-window. By combining NSW with Principal Component Analysis (PCA) and Support Vector Machine (SVM), a classification model, namely NSW-PCA-SVM, is obtained. This paper conducts experiments on three public datasets, and verifies the effectiveness of the proposed model by comparing with two basic models, i.e., SVM and PCA-SVM, and six state-of-the-art models, i.e., CDCT-WF-SVM, CDCT-2DCT-SVM, SDWT-2DWT-SVM, SDWT-WF-SVM, SDWT-2DCT-SVM and Two-Stage. The proposed approach has the following advantages in overall accuracy (OA)—take the experimental results on the Indian Pines dataset as an example: (1) Compared with SVM (OA = 53.29%) and PCA-SVM (OA = 58.44%), NSW-PCA-SVM (OA = 91.40%) effectively utilizes the spatial information of HSI and improves the classification accuracy. (2) The performance of the proposed model is mainly determined by two parameters, i.e., the window size in NSW and the number of principal components in PCA. The two parameters can be adjusted independently, making parameter adjustment more convenient. (3) When the sample size of the training set is small (20 samples per class), the proposed NSW-PCA-SVM approach achieves 2.38–18.40% advantages in OA over the six state-of-the-art models.

Texture evaluation of cooked parboiled rice using nondestructive milled whole grain near infrared spectroscopy

One consumer acceptability criterion of cooked parboiled rice is its texture, particularly hardness and toughness. The samples were obtained from parboiled rice factory for export. The hardness and toughness calibration models based on milled whole grain near infrared spectroscopy was developed. The ISO 11747 Rice-Determination of Rice Kernel Resistance to Extrusion after Cooking method was used as reference test. The models were established using partial least squares regression (PLSR), principal component regression (PCR) and support vector machine regression (SVM). The PLSR optimal calibration model of hardness with moving average smoothing pre-processing gave coefficient of determination of validation (r2), root mean square error of prediction (RMSEP) and ratio of prediction to deviation (RPD) of 0.70, 7.24 N and 1.93, respectively. The PCR optimal model of toughness using mean normalization preprocessing provided r2, RMSEP and RPD of 0.66, 38.00 Nmm and 1.75, respectively. According to RPD threshold, the models were fair for prediction application. This feasibility study suggested that the NIR protocol developed was applicable for real use due to the error of the NIR scanning and other unexplained errors was only 5% and 1% for hardness and toughness models, respectively. However, the sample preparation before texture analysis has to be improved.

Autobiography writing instruction: The journey of a teacher participating in a systemic functional linguistics genre pedagogy professional development

Teaching writing has evolved from a focus on product to an emphasis on process. More recently, interest in genre pedagogy informed by systemic functional linguistics (SFL) has emerged. This article reports on a narrative analysis of a 3rd grade teacher’s development of writing instruction. This teacher participated for 10 years in a school-university collaboration at a U.S. urban elementary school with a multilingual population. SFL genre pedagogy, instruction that draws content from SFL theory and instructional practices from the teaching and learning cycle (TLC) (Rose & Martin, 2012; Shi & Chen, 2020), successfully scaffolded writing instruction for bilingual students of different levels of English proficiency. Participation in a long term, whole school, collaborative professional development, with full principal support, allowed this teacher to evolve in the way he taught autobiography writing over the span of four years. Classroom observations revealed the steady transformation of instruction and the changes in the writing of four cohorts of bilingual students. This study adds to the extant research on the impact of SFL genre pedagogy on writing development and illustrates the value of professional development that is long term and gives agency to teachers to adapt SFL genre pedagogy to their students’ needs.

Real-time sufficient dimension reduction through principal least squares support vector machines

We propose a real-time approach for sufficient dimension reduction. Compared with popular sufficient dimension reduction methods including sliced inverse regression and principal support vector machines, the proposed principal least squares support vector machines approach enjoys better estimation of the central subspace. Furthermore, this new proposal can be used in the presence of streamed data for quick real-time updates. It is demonstrated through simulations and real data applications that our proposal performs better and faster than existing algorithms in the literature.

Powerful processing to three-dimensional facial recognition using triple information

<p><span>Face Detection plays a crucial role in identifying individuals and criminals in Security, surveillance, and footwork control systems. Face Recognition in the human is superb, and pictures can be easily identified even after years of separation. These abilities also apply to changes in a facial expression such as age, glasses, beard, or little change in the face. This method is based on 150 three-dimensional images using the Bosphorus database of a high range laser scanner in a Bogaziçi University in Turkey. This paper presents powerful processing for face recognition based on a combination of the salient information and features of the face, such as eyes and nose, for the detection of three-dimensional figures identified through analysis of surface curvature. The Trinity of the nose and two eyes were selected for applying principal component analysis algorithm and support vector machine to revealing and classification the difference between face and non-face. The results with different facial expressions and extracted from different angles have indicated the efficiency of our powerful processing.</span></p>

Rapid determination of lignocellulose in corn stover based on near-infrared reflectance spectroscopy and chemometrics methods

In this study, a rapid detection method based on near-infrared reflectance spectroscopy was proposed for measuring the contents of cellulose, hemicellulose and lignin in corn stover. In the basis of strategies of variable selection, feature extraction and nonlinear modeling, BiPLS-PCA-SVM was constructed using backward interval partial least squares combined with principal component analysis and support vector machine, which was used to improve the performance of spectral regression calibration model. For BiPLS-PCA-SVM model, the determination coefficients, root mean squared error and residual predictive deviation for the validation set were 0.906, 0.900% and 3.213 for cellulose; 0.987, 0.797% and 9.071 for hemicellulose; and 0.936, 0.264% and 4.024 for lignin, correspondingly. The results indicate that near-infrared reflectance spectroscopy combined with BiPLS-PCA-SVM can provide a reliable alternative strategy to detect contents of lignocellulosic components for pretreated corn stover in the anaerobic digestion process.

Identifying individuals with autism spectrum disorder based on the principal components of whole-brain phase synchrony

Autism spectrum disorder (ASD) is a brain disorder that develops during an early stage of childhood. Previous neuroimaging-based diagnostic models for ASD were based on static functional connectivity (FC). The nonlinear complexity of brain connectivity remains unexplored for ASD diagnosis. This study aimed to build intelligent discriminative models for ASD based on phase synchrony (PS). To this end, data from 49 patients with ASD and 41 healthy controls were obtained from the Autism Brain Imaging Data Exchange (ABIDE) project. PS between brain regions was determined using Hilbert transform. Principal component analysis (PCA) and support vector machines (SVMs) were used to build the discriminative models. PS-based models (AUC = 0.81) outperformed static FC-based models (AUC = 0.71). Furthermore, embedded functional biomarkers were discovered. Moreover, significant correlations were found between PCA-PS and the clinical severity of ASD. Together, intelligent discriminative models based on PS were established for ASD identification. The performance of the diagnostic models suggested the potential benefits of PS for clinical applications. The discriminative patterns indicated that PCA-PS features could be additional biomarkers for ASD research. Furthermore, the significant relationships between the PCA-PS features and clinical scores implied their potential use for personalized medication strategies.

Mal-Prec: computational prediction of protein Malonylation sites via machine learning based feature integration

BackgroundMalonylation is a recently discovered post-translational modification that is associated with a variety of diseases such as Type 2 Diabetes Mellitus and different types of cancers. Compared with experimental identification of malonylation sites, computational method is a time-effective process with comparatively low costs.ResultsIn this study, we proposed a novel computational model called Mal-Prec (Malonylation Prediction) for malonylation site prediction through the combination of Principal Component Analysis and Support Vector Machine. One-hot encoding, physio-chemical properties, and composition of k-spaced acid pairs were initially performed to extract sequence features. PCA was then applied to select optimal feature subsets while SVM was adopted to predict malonylation sites. Five-fold cross-validation results showed that Mal-Prec can achieve better prediction performance compared with other approaches. AUC (area under the receiver operating characteristic curves) analysis achieved 96.47 and 90.72% on 5-fold cross-validation of independent data sets, respectively.ConclusionMal-Prec is a computationally reliable method for identifying malonylation sites in protein sequences. It outperforms existing prediction tools and can serve as a useful tool for identifying and discovering novel malonylation sites in human proteins. Mal-Prec is coded in MATLAB and is publicly available at https://github.com/flyinsky6/Mal-Prec, together with the data sets used in this study.

Estimating fuel adulteration in automobiles using robust optical fiber sensors

The significance of petroleum product consumption in most countries is growing owing to different reasons like urbanization, population increase, development events and life style changes, which in turn leads to prevalent environment pollution. Adulterant refers to the substance that gets added to another and may not be legally allowed in most of the cases. A petroleum fuel is one such case which is vulnerable to adulteration specifically for improving the profit margins. Detection of this petroleum fuels adulteration is challenging as they are naturally present in the compounds already. The compositional variations of these fuels are determined using various physico-chemical properties measurements. For discriminating the adulterated samples from the unaltered ones, the statistical designs along with the data mining help. Monitoring of the quality of fuel is essential at the distribution point for the prevention of adulteration. We propose to use a fuel adulterations setup that is portable, in expensive and is capable of providing the results in a short time. This includes the use of a light weight optical fiber sensor that gives high performance with low attenuation and there are no fire hazards, as well as they are resistant to harsh environments for testing. The distilled curves along with principal component analysis and support vector machine based classification helps us to build a model that is capable of this adulteration detection.

Guiding Principals: Middle-Manager Coaching and Human-Capital Reform

Background/Context School leaders are central to state and district human-capital reforms (HCRs), yet they are rarely equipped with the skills to implement new evaluation, professional development, and personnel data systems. Although districts increasingly offer principals coaching and training, there has been limited empirical work on how these supports influence principals’ HCR-related practices. Purpose Drawing on a two-year, mixed-methods study in the Los Angeles Unified School District (LAUSD), this article examines the role of principal supervisors in HCRs. We ask: What role did principal supervisors (Instructional Directors [IDs]) play in the implementation of human-capital reforms? What did high-quality coaching on the part of IDs look like in this context? Research Design Our two-part analysis draws upon survey and interview data. First, we conducted descriptive analyses and significance testing using principal and ID survey data to examine the correlations among principals’ ratings of ID coaching quality, ID coaching practices, and principals’ implementation of HCRs. Second, we conducted in-depth interviews, using a think-aloud protocol, with two sets of IDs—those consistently highly-rated and those with mixed ratings—who were identified using principals’ reports of coaching quality. Following interview coding, we created various case-ordered metamatrix displays to analyze our qualitative data in order to identify patterns in coaching strategy and approach across IDs, content, and contexts. Findings First, our survey data indicate that receiving high-quality coaching from IDs is correlated with stronger principal support for and implementation of HCRs. Our survey findings further illustrate that IDs support a wide range of principals’ HCR activities. Second, our think-aloud interviews with case IDs demonstrate that coaching strategy and approach vary between consistently highly-rated and mixed-rated coaches: Consistently highly-rated IDs emphasize the importance of engaging in, or defining HCR problems as, joint work alongside principals, while mixed-rated IDs often emphasize the use of tools to guide principal improvement. We find that, on the whole, the consistently highly-rated IDs in our sample employ a nondirective approach to coaching more often than mixed-rated coaches. Conclusions These findings contribute to a growing literature on the crucial role of principal supervisors as coaches to improve principals’ instructional leadership and policy implementation. While exploratory, this study offers the first steps toward building greater evidence of the connections between high-quality coaching and policy implementation, and it may have implications for the design and implementation of professional development for principal supervisors and the selection and placement of supervisors with principals.

A Comprehensive Method for Assessing Meat Freshness Using Fusing Electronic Nose, Computer Vision, and Artificial Tactile Technologies

The traditional methods cannot be used to meet the requirements of rapid and objective detection of meat freshness. Electronic nose (E-Nose), computer vision (CV), and artificial tactile (AT) sensory technologies can be used to mimic humans’ compressive sensory functions of smell, look, and touch when making judgement of meat quality (freshness). Though individual E-Nose, CV, and AT sensory technologies have been used to detect the meat freshness, the detection results vary and are not reliable. In this paper, a new method has been proposed through the integration of E-Nose, CV, and AT sensory technologies for capturing comprehensive meat freshness parameters and the data fusion method for analysing the complicated data with different dimensions and units of six odour parameters of E-Nose, 9 colour parameters of CV, and 4 rubbery parameters of AT for effective meat freshness detection. The pork and chicken meats have been selected for a validation test. The total volatile base nitrogen (TVB-N) assays are used to define meat freshness as the standard criteria for validating the effectiveness of the proposed method. The principal component analysis (PCA) and support vector machine (SVM) are used as unsupervised and supervised pattern recognition methods to analyse the source data and the fusion data of the three instruments, respectively. The experimental and data analysis results show that compared to a single technology, the fusion of E-Nose, CV, and AT technologies significantly improves the detection performance of various freshness meat products. In addition, partial least squares (PLS) is used to construct TVB-N value prediction models, in which the fusion data is input. The root mean square error predictions (RMSEP) for the sample pork and chicken meats are 1.21 and 0.98, respectively, in which the coefficient of determination (R2) is 0.91 and 0.94. This means that the proposed method can be used to effectively detect meat freshness and the storage time (days).

Experimental evaluation of rock disintegration detection in drilling by a new acoustic sensor method

A simple and reliable method for monitoring the rock properties in drilling is recognized as a global challenge, especially for cluster and infill development well drilling in oil and gas production. To improve the existing limitations in the detection of formation lithology in drilling systems, an acoustic sensor approach for rock disintegration in the drilling using a polycrystalline diamond compact (PDC) drill bit is developed and evaluated. In this paper, we report the application of a special acoustic sensor device to qualitatively survey rock disintegration by rotary drilling. In this work, an acoustic focusing component is designed to assist a condenser microphone in detecting signals emerging in the process of drilling as an integrating information resource. A time-frequency with zoom fast-Fourier-transform analysis method is presented to recognize the acoustic signal features in breaking different rocks. Additionally, principal component analysis and Support Vector Machine method are used to check the qualitative analysis results for the monitored rock information. An experimental investigation considered four types of rock broken by a PDC drill bit with drilling weights ranging from 5 to 12.5 kN and drilling rates ranging from 64 to 128 r/min. To verify the reliability and accuracy of our acoustic rock detection method, two sensors installed in the orthogonal direction are examined. Furthermore, the coherent analysis was used to reveal the correlation of the output auto-correlation spectrum for each measurement. The results show that there is a good average percentage of lithology calculated by the time-frequency analysis results between the acoustic signal features and different types of rock. Accordingly, the acoustic sensor approach can help enhance the detection of rock properties in drilling systems, laying the foundation for quantitative lithology identification in more complex drilling.

Estimating leaf chlorophyll content by laser-induced fluorescence technology at different viewing zenith angles.

Leaf chlorophyll content (LCC) is a key indicator of a plant's physiological status. Fast and non-destructive monitoring of chlorophyll content in plants through remote sensing is very important for accurate diagnosis and assessment of plant growth. Through the use of laser-induced fluorescence (LIF) technology, this study aims to compare the predictive ability of different single fluorescence characteristic and fluorescence characteristic combinations at various viewing zenith angles (VZAs) combined with multivariate analysis algorithms, such as principal component analysis (PCA) and support vector machine (SVM), for estimating the LCC of plants. The SVM models of LCC estimation were proposed, and fluorescence characteristics-fluorescence peak (FP), fluorescence ratio (FR), PCA, and first-derivative (FD) parameter-and fluorescence characteristic combinations (FP+FR, FP+FD, FR+FD, FP+FR+FD) were used as input variables for the models. Experimental results demonstrated that the effect of single fluorescence characteristics on the predictive performance of SVM models was: FR>FD>FP>PCA. Compared with other models, 0° SVM was the optimal model for estimating LCC by higher R2. The fluorescence spectra and FD spectra observed at 0° and 30° were superior to those observed at 15°, 45°, and 60°. Thus, appropriate VZA must also be considered, as it can improve the accuracy of LCC monitoring. In addition, compared with single fluorescence characteristic, the FP+FR+FD was the optimal combination of fluorescence characteristics to estimate the LCC for the SVM model by higher R2, indicating better predictive performance. The experimental results show that the combination of LIF technology and multivariate analysis can be effectively used for LCC monitoring and has broad development prospects.

Australian Primary School Principals', Teachers', and Parents' Attitudes and Barriers to Changing School Uniform Policies From Traditional Uniforms to Sports Uniforms.

To determine Australian primary school principals', teachers', and parents' attitudes to changing school uniform policies to allow students to wear sports uniforms every day and to assess associations between participant characteristics and their attitudes. A secondary aim was to identify principals' and teachers' perceived barriers to uniform changes. Cross-sectional surveys of principals, teachers, and parents of children in grades 2 to 3 (age 7-10y) from 62 Australian primary schools (Oct 2017-Mar 2018) were undertaken. Mixed logistic regression analyses assessed the associations between participant characteristics and attitudes toward uniform changes. In total, 73% of the principals (38/52) who responded reported that their school only allowed children to wear a sports uniform on sports days. Overall, 38% of the principals (18/47), 63% of the teachers (334/579), and 78% of the parents (965/1231) reported they would support a policy that allowed children to wear daily sports uniforms. The most commonly reported barrier was the perception that sports uniforms were not appropriate for formal occasions. Although the majority of the principals were not supportive of a change to a daily sports uniform, the majority of the teachers and parents were. Strategies to improve principal support may be required if broader adoption of physical activity-supporting uniforms is to be achieved.

Electronic Nose Testing for Confined Space Application Utilizes Principal Component Analysis and Support Vector Machine

A confined space has a limited space for entry and exit but it is large enough for workers to enter and perform work inside. It is not designed for continuous occupancy because it can contribute atmospheric hazards accidents that threaten the worker safety and industry progress. In this work, we reported the testing an instrument to assist workers for atmosphere testing during pre-entry. An electronic nose (e-nose) using specific sensor arrays is the integration between hardware and software that able to sense different concentrations of gases in an air sample using pattern recognition techniques. The instrument utilizes multivariate statistical analysis which is Principal Component Analysis (PCA) for discriminate the different concentrations of gases and the Support Vector Machine (SVM) to classify the acquired data from the air sample. The instrument was successfully tested using diesel, gasoline, petrol and thinner. The results show that the instrument able to discriminate an air sample using PCA with total variation for 99.94%, while the classifier success rate for SVM indicates at 98.21% for train performance and 95.83% for test performance. This will contribute significantly to acquiring a new and alternative method of using the instrument for monitoring the atmospheric hazards in confined space to ensure the safety of workers during work progress in a confined space.

Rapid and non-destructive seed viability prediction using near-infrared hyperspectral imaging coupled with a deep learning approach

Seeds are the basis of the agricultural food industry, greater insights into seed viability before sowing could improve storage management and field performance. In the present study, we aimed to address this issue by using highly cost-efficient near-infrared hyperspectral imaging (NIR-HSI) and a convolutional neural network (CNN) deep learning approach. An NIR-HSI camera was used because it can recognize both molecular vibration information (i.e. chemical component differences) and its spatial distribution in each seed sample; this camera is much more informative than a regular RGB digital camera. Using this technology, the emphasis of this study was firstly to provide a methodology for enhancing the interpretability of viable and non-viable seeds via principal component analysis (PCA) and support vector machine (SVM) viability classification analysis of NIR-HSI data. A CNN was then constructed to“cognize” the differences in viable and non-inviable seeds and classify them automatically. Experimental results indicate that the methodology produces a ~90% classification accuracy for both a five-fold cross-validation set and a test set of naturally aged Japanese mustard spinach seeds. Therefore, this study provides a new strategy for effective and practical seed viability prediction.