Advanced detection tools in food fraud: A systematic review for holistic and rational detection method based on research and patents.

Modern food chain supply management necessitates the dire need for mitigating food fraud and adulterations. This holistic review addresses different advanced detection technologies coupled with chemometrics to identify various types of adulterated foods. The data on research, patent and systematic review analyses (2018-2023) revealed both destructive and non-destructive methods to demarcate a rational approach for food fraud detection in various countries. These intricate hygiene standards and AI-based technology are also summarized for further prospective research. Chemometrics or AI-based techniques for extensive food fraud detection are demanded. A systematic assessment reveals that various methods to detect food fraud involving multiple substances need to be simple, expeditious, precise, cost-effective, eco-friendly and non-intrusive. The scrutiny resulted in 39 relevant experimental data sets answering key questions. However, additional research is necessitated for an affirmative conclusion in food fraud detection system with modern AI and machine learning approaches.

A novel method for detecting credit card fraud problems.

Credit card fraud is a significant problem that costs billions of dollars annually. Detecting fraudulent transactions is challenging due to the imbalance in class distribution, where the majority of transactions are legitimate. While pre-processing techniques such as oversampling of minority classes are commonly used to address this issue, they often generate unrealistic or overgeneralized samples. This paper proposes a method called autoencoder with probabilistic xgboost based on SMOTE and CGAN(AE-XGB-SMOTE-CGAN) for detecting credit card frauds.AE-XGB-SMOTE-CGAN is a novel method proposed for credit card fraud detection problems. The credit card fraud dataset comes from a real dataset anonymized by a bank and is highly imbalanced, with normal data far greater than fraud data. Autoencoder (AE) is used to extract relevant features from the dataset, enhancing the ability of feature representation learning, and are then fed into xgboost for classification according to the threshold. Additionally, in this study, we propose a novel approach that hybridizes Generative Adversarial Network (GAN) and Synthetic Minority Over-Sampling Technique (SMOTE) to tackle class imbalance problems. Our two-phase oversampling approach involves knowledge transfer and leverages the synergies of SMOTE and GAN. Specifically, GAN transforms the unrealistic or overgeneralized samples generated by SMOTE into realistic data distributions where there is not enough minority class data available for GAN to process effectively on its own. SMOTE is used to address class imbalance issues and CGAN is used to generate new, realistic data to supplement the original dataset. The AE-XGB-SMOTE-CGAN algorithm is also compared to other commonly used machine learning algorithms, such as KNN and Light GBM, and shows an overall improvement of 2% in terms of the ACC index compared to these algorithms. The AE-XGB-SMOTE-CGAN algorithm also outperforms KNN in terms of the MCC index by 30% when the threshold is set to 0.35. This indicates that the AE-XGB-SMOTE-CGAN algorithm has higher accuracy, true positive rate, true negative rate, and Matthew's correlation coefficient, making it a promising method for detecting credit card fraud.

FraudAuditor: A Visual Analytics Approach for Collusive Fraud in Health Insurance.

Collusive fraud, in which multiple fraudsters collude to defraud health insurance funds, threatens the operation of the healthcare system. However, existing statistical and machine learning-based methods have limited ability to detect fraud in the scenario of health insurance due to the high similarity of fraudulent behaviors to normal medical visits and the lack of labeled data. To ensure the accuracy of the detection results, expert knowledge needs to be integrated with the fraud detection process. By working closely with health insurance audit experts, we propose FraudAuditor, a three-stage visual analytics approach to collusive fraud detection in health insurance. Specifically, we first allow users to interactively construct a co-visit network to holistically model the visit relationships of different patients. Second, an improved community detection algorithm that considers the strength of fraud likelihood is designed to detect suspicious fraudulent groups. Finally, through our visual interface, users can compare, investigate, and verify suspicious patient behavior with tailored visualizations that support different time scales. We conducted case studies in a real-world healthcare scenario, i.e., to help locate the actual fraud group and exclude the false positive group. The results and expert feedback proved the effectiveness and usability of the approach.

An Affordable NIR Spectroscopic System for Fraud Detection in Olive Oil.

Adulterations of olive oil are performed by adding seed oils to this high-quality product, which are cheaper than olive oils. Food safety controls have been established by the European Union to avoid these episodes. Most of these methodologies require expensive equipment, time-consuming procedures, and expert personnel to execute. Near-infrared spectroscopy (NIRS) technology has many applications in the food processing industry. It analyzes food safety and quality parameters along the food chain. Using principal component analysis (PCA), the differences and similarities between olive oil and seed oils (sesame, sunflower, and flax oil) have been evaluated. To quantify the percentage of adulterated seed oil in olive oils, partial least squares (PLS) have been employed. A total of 96 samples of olive oil adulterated with seed oils were prepared. These samples were used to build a spectra library covering various mixtures containing seed oils and olive oil contents. Eighteen chemometric models were developed by combining the first and second derivatives with Standard Normal Variable (SNV) for scatter correction to classify and quantify seed oil adulteration and percentage. The results obtained for all seed oils show excellent coefficients of determination for calibration higher than 0.80. Because the instrumental aspects are not generally sufficiently addressed in the articles, we include a specific section on some key aspects of developing a high-performance and cost-effective NIR spectroscopy solution for fraud detection in olive oil. First, spectroscopy architectures are introduced, especially the Texas Instruments Digital Light Processing (DLP) technology for spectroscopy that has been used in this work. These results demonstrate that the portable prototype can be used as an effective tool to detect food fraud in liquid samples.

Rapid detection of sesame oil multiple adulteration using a portable Raman spectrometer.

Multiple adulteration is a commonly used fraud of illegal traders to mask the traditional adulteration detection methods. In this study, rapid detection of multiple adulteration of sesame oil was proposed using a portable Raman spectrometer. Two strategies including simplex theory of mixtures and D-optimal mixture design were used to conduct variable selection and model evaluation, respectively. Based on simplex theory of mixtures, the important variables were selected by orthogonal partial least squares discriminant analysis of preprocessed Raman spectra of sesame oils and four adulterant oils. Moreover, multiple adulteration identification model was built by one-class partial least squares and validated by representative adulterated samples prepared by D-Optimal mixture design. The validation results show that 40 sesame oils adulterated with four types of adulterant oils can be correctly identified, indicating Raman spectroscopy is an effective tool for the detection of multiple adulteration of sesame oil, especially for on-site applications.

Nuclear magnetic resonance spectroscopy in extra virgin olive oil authentication.

Extra virgin olive oil (EVOO) is a high-quality product that has become one of the stars in the food fraud context in recent years. EVOO can encounter different types of fraud, from adulteration with cheaper oils to mislabeling, and for this reason, the assessment of its authenticity and traceability can be challenging. There are several officially recognized analytical methods for its authentication, but they are not able to unambiguously trace the geographical and botanical origin of EVOOs. The application of nuclear magnetic resonance (NMR) spectroscopy to EVOO is reviewed here as a reliable and rapid tool to verify different aspects of its adulteration, such as undeclared blends with cheaper oils and cultivar and geographical origin mislabeling. This technique makes it possible to use both targeted and untargeted approaches and to determine the olive oil metabolomic profile and the quantification of its constituents.

One-step ultra-sensitive immunochromatographic strip authenticating an emergent fraud acetophenetidin in herbal tea.

Herbal tea is a highly popular and widely consumed beverage. However, a pain-relieving and fever-reducing drug, acetophenetidin, was recently found to illegally occur in herbal tea for a fraud purpose. Due to the potential health risk and urgent requirement for on-site screening method, a one-step and high specificity strip for identifying acetophenetidin was developed for the first time. Assisted by computational chemistry, four haptens were designed to prepare immunogens and coating antigens for antibody generation, and a specific antibody with ultra-sensitivity and high specificity was generated, showing half maximal inhibitory (IC50) of 16.46 ng/mL for acetophenetidin, less than 3.5% of cross-reactivity to analogs by ELISA. A gold nanoparticles immunochromatographic strip was developed for detection of acetophenetidin in herbal tea, demonstrating a cut-off value of 160 ng/mL and a limit of detection of 1.63 ng/mL. The recovery rates were ranged from 102.2% to 106.1% with coefficient of variation between 2.21% and 7.20%. The analysis of real samples (n = 20) by the strip was well correlated with that of the confirmatory method, liquid chromatography-tandem mass spectrometry. The proposed strip has the potential to be used for rapid screening of acetophenetidin in herbal tea.

Fraud detection in crude palm oil using SERS combined with chemometrics.

Edible crude palm oil (CPO) is a vital oil utilized in various industries, including food, pharmaceuticals, and domestic cooking. Unfortunately, reports of CPO adulteration with harmful Sudan dyes have surfaced over the years. Surface-enhanced Raman spectroscopy (SERS) and chemometrics were employed to detect Sudan dyes adulteration in CPO within 900 - 1800 cm- 1 Raman peak. The concentration of Sudan dyes detected in CPO samples ranged between 0.005 and 4 ppm. The principal component analysis (PCA) model detected Sudan II and Sudan IV in CPO with 99.88 and 99.90% accuracy. Linear discriminant analysis (LDA) and K-Nearest Neighbors (KNN) also recorded high detection rates of Sudan II and IV dyes in CPO. Sudan II and IV dyes could be detected at 0.0028 ppm and 0.0019 ppm by this sensor. The performance of the Au@Ag SERS sensor was comparable to that of HPLC. This study proved SERS and chemometrics can be used to authenticate edible CPO.

Wading through Molasses: A qualitative examination of the experiences, perceptions, attitudes, and knowledge of Australian medical practitioners regarding medical billing.

BACKGROUND: Medical billing errors and fraud have been described as one of the last "great unreduced healthcare costs," with some commentators suggesting measurable average losses from this phenomenon are 7% of total health expenditure. In Australia, it has been estimated that leakage from Medicare caused by non-compliant medical billing may be 10-15% of the scheme's total cost. Despite a growing body of international research, mostly from the U.S, suggesting that rather than deliberately abusing the health financing systems they operate within, medical practitioners may be struggling to understand complex and highly interpretive medical billing rules, there is a lack of research in this area in Australia. The aim of this study was to address this research gap by examining the experiences of medical practitioners through the first qualitative study undertaken in Australia, which may have relevance in multiple jurisdictions. METHOD: This study interviewed 27 specialist and general medical practitioners who claim Medicare reimbursements in their daily practice. Interviews were recorded, transcribed, and analysed using thematic analysis. RESULTS: The qualitative data revealed five themes including inadequate induction, poor legal literacy, absence of reliable advice and support, fear and deference, and unmet opportunities for improvement. CONCLUSION: The qualitative data presented in this study suggest Australian medical practitioners are ill-equipped to manage their Medicare compliance obligations, have low levels of legal literacy and desire education, clarity and certainty around complex billing standards and rules. Non-compliant medical billing under Australia's Medicare scheme is a nuanced phenomenon that may be far more complex than previously thought and learnings from this study may offer important insights for other countries seeking solutions to the phenomenon of health system leakage. Strategies to address the barriers and deficiencies identified by participants in this study will require a multi-pronged approach. The data suggest that the current punitive system of ensuring compliance by Australian medical practitioners is not fit for purpose.

Multifunctional Probes with High Utilization Rates: Self-Assembled Merocyanine Nanoparticles in Water as Acid-Base Indicators and Mitochondrion-Targeting Chemotherapeutic Agents.

Multifunctional probes with high utilization rates have great value in practical applications in various fields such as cancer diagnosis and therapy. Here we have synthesized two organic molecules based on merocyanine. They can self-assemble in water to form ∼1.5 nm nanoparticles. Both of them have good application potential in fluorescent anticounterfeit printing ink and pH detection. More importantly, they have excellent mitochondrial targeting ability, intracellular red light and near-infrared dual-channel imaging ability, strong antiphotobleaching ability, and in vivo and in vitro near-infrared imaging capabilities, showing superior chemotherapy capabilities and biocompatibility in the 4T1 tumor-bearing mouse model.

Fraud Detection of Herbal Medicines Based on Modern Analytical Technologies Combine with Chemometrics Approach: A Review.

Fraud in herbal medicines (HMs), commonplace throughout human history, is significantly related to medicinal effects with sometimes lethal consequences. Major HMs fraud events seem to occur with a certain regularity, such as substitution by counterfeits, adulteration by addition of inferior production-own materials, adulteration by chemical compounds, and adulteration by addition of foreign matter. The assessment of HMs fraud is in urgent demand to guarantee consumer protection against the four fraudulent activities. In this review, three analysis platforms (targeted, non-targeted, and the combination of non-targeted and targeted analysis) were introduced and summarized. Furthermore, the integration of analysis technology and chemometrics method (e.g., class-modeling, discrimination, and regression method) have also been discussed. Each integration shows different applicability depending on their advantages, drawbacks, and some factors, such as the explicit objective analysis or the nature of four types of HMs fraud. In an attempt to better solve four typical HMs fraud, appropriate analytical strategies are advised and illustrated with several typical studies. The article provides a general workflow of analysis methods that have been used for detection of HMs fraud. All analysis technologies and chemometrics methods applied can conduce to excellent reference value for further exploration of analysis methods in HMs fraud.

Three centuries on the science of coffee authenticity control.

Food authenticity relies on genuineness and reliability according to the information displayed on the package. Since the 18th century, when coffee became popularized in the West, adulteration began. Several methods have been developed to detect different kinds of frauds and they have evolved as demands increased and new technologies were introduced. The evolution of the science of coffee authenticity control in the past three centuries is reviewed, focusing on the discrimination between coffee and other foods or between coffee and its by-products. The earliest chemical and physical methods are presented followed by methods developed in the 20th and 21st centuries using microscopy, chromatography and spectroscopy associated with advanced statistical tools, and DNA-based methods. In addition to non-food material, before the 20th century, chicory was the most studied food-adulterant. From the 20th century on, corn, coffee by-products, and barley were the most studied, followed by chicory, rice and other food items. Matrix effects seem to be among the most challenging problems in these approaches, associated with variations in roast degree, particle size (particularly in spectroscopy-based methods), and lack of control over reference samples regarding species and purity. Limits of detection vary considerably within each category, with most limits being too high for commercial use. DNA-based methods appear to be promising to assess coffee authenticity, given that the limits of detection and quantification are low, and specificity is higher than in other methods. Nevertheless, as roast intensity increases, the sensitivity of the method decreases. So far, most reported methods have not been validated and only a few have been tested on commercial brands, except for those involving microscopy which has been highly used for monitoring coffee authenticity although not always efficiently enough.

Pharmacobotanical characterization of Monteverdia ilicifolia (Mart. ex Reiss.) Biral leaves and its adulterants sold as medicinal tea in Brazil: a contribution to quality control / Caracterización farmacobotánica de hojas de Monteverda ilicifolia (Mart. ex Reiss.) Biral y sus adulterantes vendidos como té medicinal en Brasil: una contribución al contol de calidad

Leaves of Monteverdia ilicifolia ("espinheira-santa") are considered a medicinal tea by the Brazilian Sanitary Surveillance Agency (Anvisa), by their anti-dyspeptic, anti-acid and protective of the gastric mucosa properties. Their spiny margins are similar to those of other botanical species, which may lead to misidentifications. The aim of this work was to evaluate the authenticity of 32 samples of herbal drugs commercialized as "espinheira-santa" in the formal trade in Brazil, by macro and microscopic morphological studies of the leaves. The evaluation of the botanical authenticity was based on leaf venation patterns, shape and anatomy of the petiole and midrib region in cross section, vascular system arrangement and epidermal characters. Analysis of these characters compared to literature data suggests that 34% of the samples are M. ilicifolia and the remaining 66% are Sorocea bonplandii, a species with no clinical studies assuring its effective and safe use, representing thus a potential risk to public health.

Las hojas de Monteverdia ilicifolia ("espinheira-santa") son consideradas tés medicinales por la Agencia Nacional de Vigilancia Sanitaria (Anvisa), Brasil, por las indicaciones terapéuticas como antidispépticos, antiácidos y protectores de la mucosa gástrica. Sus márgenes foliares espinescentes se parecen a los de otras especies, conduciendo a identificaciones erróneas. El objetivo de este trabajo fue evaluar la autenticidad de 32 muestras de drogas vegetales vendidas como "espinheira-santa" en el comercio legal de Brasil, realizando un estudio morfológico de las hojas. Esta evaluación se ha basado en observar de la hoja, sus patrones de nerviación, su forma y anatomía (pecíolo y nervadura central en corte transversal), disposición del sistema vascular y caracteres epidérmicos. El análisis de la morfología, comparado con los datos de literatura, sugiere que el 34% de las muestras son M. ilicifoliay el 66% son Sorocea bonplandii, una especie que no cuenta con los estudios clínicos que garantizan su uso efectivo y seguro, representando un riesgo para la salud pública.

[Fraud in nutritional supplements for athletes: a narrative review]. / Fraude en suplementos nutricionales para deportistas: revisión narrativa.

INTRODUCTION: The use of nutritional supplements (NSs) is a widespread practice in both professional and amateur athletes. Their choice must be based on criteria of safety, legality, and effectiveness. There are several risks that some of these products can trigger, this being a reason for fraud to the consumer. By way of example, the presence of a legal substance in doses higher than its recommendation may cause unwanted side effects; on the other hand, its presence in quantities lower than necessary (effective threshold) may not have the effect expected or announced by the product. Therefore, the objective of this review was to describe the existence of fraud in relation to nutritional labeling and/or to the content of NSs intended for athletes. A narrative review was conducted through PubMed in which 16 studies met the inclusion criteria. The results show that nutritional labeling differs from the actual amounts of ingredients. This type of fraud has been found in protein powder supplements, pre-workout mixtures, creatine, and beverages containing vitamins and minerals, among others. Fraud due to errors in the labeling by omission of substances present in the product or by errors in the analysis or declaration of quantities, is high. This lack of transparency in the actual composition of labels can contribute to sports performance and/or health problems in the athlete. Therefore, an improvement of the legislation on NS is necessary to ensure their safety, effectiveness, and legality.

INTRODUCCIÓN: El uso de suplementos nutricionales (SN) es una práctica muy extendida tanto entre los deportistas profesionales como entre los amateurs. Su elección debe realizarse basándose en criterios de seguridad, legalidad y eficacia. Son varios los riesgos que algunos de estos productos pueden desencadenar, siendo motivo de fraude para el consumidor. A modo de ejemplo, la presencia de una sustancia legal en dosis superiores a sus recomendaciones puede ocasionar efectos secundarios indeseados; por el contrario, su presencia en cantidades inferiores a las necesarias (umbral efectivo) puede no provocar el efecto esperado o anunciado por el producto. Por ello, el objetivo de la presente revisión es describir la existencia de fraude en relación con el etiquetado nutricional y/o el contenido de los SN destinados a deportistas. Se realizó una revisión narrativa a través de Pubmed en la que 16 estudios cumplieron los criterios de inclusión. Los resultados muestran que el etiquetado nutricional difiere de la cantidad real de ingredientes. Este tipo de fraude se ha encontrado en suplementos de proteína en polvo, mezclas preentrenamiento, creatina y bebidas que contienen vitaminas y minerales, entre otros. El fraude debido a errores en el etiquetado, por omisión de sustancias presentes en el producto o por errores en el análisis o declaración de las cantidades, es alto. Esta falta de transparencia en la composición real del etiquetado puede contribuir a problemas de rendimiento deportivo y/o de salud en el deportista. Por lo tanto, es necesaria una mejora de la legislación de los SN para garantizar su seguridad, eficacia y legalidad.

Multispecies Identification of Oilseed- and Meat-Specific Proteins and Heat-Stable Peptide Markers in Food Products.

Consumer demand for both plant products and meat products enriched with plant raw materials is constantly increasing. Therefore, new versatile and reliable methods are needed to find and combat fraudulent practices in processed foods. The objective of this study was to identify oilseed species-specific peptide markers and meat-specific markers that were resistant to processing, for multispecies authentication of different meat and vegan food products using the proteomic LC-MS/MS method. To assess the limit of detection (LOD) for hemp proteins, cooked meatballs consisting of three meat species and hemp cake at a final concentration of up to 7.4% were examined. Hemp addition at a low concentration of below 1% was detected. The LOD for edestin subunits and albumin was 0.9% (w/w), whereas for 7S vicilin-like protein it was 4.2% (w/w). Specific heat-stable peptides unique to hemp seeds, flaxseed, nigella, pumpkin, sesame, and sunflower seeds, as well as guinea fowl, rabbit, pork, and chicken meat, were detected in different meat and vegan foods. Most of the oilseed-specific peptides were identified as processing-resistant markers belonging to 11S globulin subunits, namely conlinin, edestin, helianthinin, pumpkin vicilin-like or late embryogenesis proteins, and sesame legumin-like as well as 2S albumins and oleosin isoforms or selected enzymic proteins.

Detection of camellia oil adulteration using chemometrics based on fatty acids GC fingerprints and phytosterols GC-MS fingerprints.

The fatty acid, squalene, and phytosterols, coupled to chemometrics were utilized to detect the adulteration of camellia oil (CAO) with palm superolein (PAO), refined olive oil (ROO), high oleic- sunflower oil (HO-SUO), sunflower oil (SUO), corn oil (COO), rice bran oil (RBO), rice oil (RIO), peanut oil (PEO), sesame oil (SEO), soybean oil (SOO), and rapeseed oil (RAO). CAO was characterized with higher triterpene alcohols, thus differentiated from other vegetable oils in principle component analysis (PCA). Using partial least squares-discriminant analysis (PLS-DA), CAO adulterated with PAO, ROO, HO-SUO, SUO, COO, RBO, RIO, PEO, SEO, SOO, RAO (5%-100%, w/w), could be classified, especially higher than 92.31% of the total discrimination accuracy, at an adulterated ratio above 30%. With less than 22 potential key markers selected by the variable importance in projection (VIP), the optimized PLS models were confirmed to be accurate for the adulterated level prediction in CAO.

Recent advances in analytical techniques for the detection of adulteration and authenticity of bee products - A review.

Bee products have been considered as functional foods for a long time in China because of their wide range of biological activity. China has the largest number of bee colonies and the highest production of bee products in the world. Major bee products include honey, royal jelly, propolis and bee pollen. In recent years, consumption of bee products in China has been increasing due to an increased public awareness of their nutritional and health benefits. With the development of the Chinese economy and the improvement of people's living standards, high-end and gift-oriented products have become more popular and bee products are one of the options. However, the production of bee products cannot increase rapidly in short term and this is a driver for substantial economic-motivated adulteration. This is compounded by globalisation of supply chains which has also resulted in a rise in bee products fraud. These illicit products are eroding market prices and consumer trust, causing significant damage to the beekeeping industry. In order to provide information or solutions for regulators and consumers, in this article, we review he characteristics of bee products in China and the current situation regarding adulteration and authenticity of bee products. Moreover, advances in analytical techniques for detection of adulteration and authenticity of bee products including sensory techniques, DNA methods, isotope ratio mass spectrometry, spectroscopic techniques and mass spectrometry are reviewed. Finally, the applications and limitations of analytical methods in authentication are critically assessed. Suggestions are also put forward for the future management of China's bee products industry.

HPLC-DAD fingerprints combined with chemometric techniques for the authentication of plucking seasons of Laoshan green tea.

Laoshan green teas plucked in summer and autumn were measured by high performance liquid chromatography-diode array detector (HPLC-DAD). After baseline correction, the fingerprints data were resolved by multivariate curve resolution-alternating least squares (MCR-ALS) and a total of 57 components were acquired. Relative concentrations of these components were afterwards applied to distinguish plucking seasons using principal component analysis (PCA), support vector machines (SVM) and partial least squares-discriminant analysis (PLS-DA). For both SVM and PLS-DA models, the total recognition rates of training set, cross-validation and testing set were 100%, 91.3% and 100%, respectively. Besides, three variable selection methods were employed to determine characteristic components for the authentication of summer and autumn teas. Results showed that PLS-DA model based on three characteristic components selected by VIP possesses identical predictive ability as the original model. This study demonstrated that our proposed strategy is competent for the authentication of plucking seasons of Laoshan green tea.

Exploration of total synchronous fluorescence spectroscopy combined with pre-trained convolutional neural network in the identification and quantification of vegetable oil.

In order to distinguish different vegetable oils, adulterated vegetable oils, and to identify and quantify counterfeit vegetable oils, a method based on a small sample size of total synchronous fluorescence (TSyF) spectra combined with convolutional neural network (CNN) was proposed. Four typical vegetable oils were classified by three ways of fine-tuning the pre-trained CNN, the pre-trained CNN as a feature extractor, and traditional chemometrics. The pre-trained CNN was combined with support vector machines to distinguish adulterated sesame oil and counterfeit sesame oil separately with 100% correct classification rates. The pre-trained CNN combined with partial least square regression was used to predict the level of counterfeit sesame oil. The coefficient of determination for calibration (Rc2) values were all greater than 0.99, and the root mean square errors of validation were 0.81% and 1.72%, respectively. These results show that it is feasible to combine TSyF spectra with CNN for vegetable oil identification.