Academic literature on the topic 'Odor quality prediction'

Mapping molecular structure to odor perception is a key challenge in olfaction. We used graph neural networks to generate a principal odor map (POM) that preserves perceptual relationships and enables odor quality prediction for previously uncharacterized odorants. The model was as reliable as a human in describing odor quality: On a prospective validation set of 400 out-of-sample odorants, the model-generated odor profile more closely matched the trained panel mean than did the median panelist. By applying simple, interpretable, theoretically rooted transformations, the POM outperformed chemoinformatic models on several other odor prediction tasks, indicating that the POM successfully encoded a generalized map of structure-odor relationships. This approach broadly enables odor prediction and paves the way toward digitizing odors.

The odor emitted from a wastewater treatment plant (WWTP) is an important environmental problem. An estimation of odor emission rate is difficult to detect and quantify. To address this, various approaches including the development of emission factors and measurement using a closed chamber have been employed. However, the evaluation of odor emission involves huge manpower, time, and cost. An artificial neural network (ANN) is recognized as an efficient method to find correlations between nonlinear data and prediction of future data based on these correlations. Due to its usefulness, ANN is used to solve complicated problems in various disciplines of sciences and engineering. In this study, a method to predict the odor concentration in a WWTP using ANN was developed. The odor concentration emitted from a WWTP was predicted by the ANN based on water quality data such as biological oxygen demand, dissolved oxygen, and pH. The water quality and odor concentration data from the WWTP were measured seasonally in spring, summer, and autumn and these were used as input variations to the ANN model. The odor predicted by the ANN model was compared with the measured data and the prediction accuracy was estimated. Suggestions for improving prediction accuracy are presented.

Black-odor rivers are polluted urban rivers that often are black in color and emit a foul odor. They are a severe problem in aquatic systems because they can negatively impact the living conditions of residents and the functioning of ecosystems and local economies. Therefore, it is crucial to identify ways to mitigate the water quality parameters that characterize black-odor rivers. In this study, we tested the efficacy of an electronic nose (E-nose), which was inexpensive, fast, and easy to operate, for qualitative recognition analysis and quantitative parameter prediction of samples collected from the Yueliang River in Huzhou City. The E-nose sensors were cross-sensitive to the volatile compounds in black-odor water. The device recognized the samples from different river sites with 100% accuracy based on linear discriminant analysis. For water quality parameter predictions, partial least squares regression models based on E-nose signals were established, and the coefficients between the actual water quality parameters (pH, chemical oxygen demand, total nitrogen content, and total phosphorous content) and the predicted values were very high (R2 > 0.90) both in the training and testing sets. These results indicate that E-nose technology can be a fast, easy-to-build, and cost-effective detection system for black-odor river monitoring.

Milk quality prediction is considered a vital research area due to increase the need for obtain sustainable development goals. This study aims to predict milk quality by integrate gated recurrent units (GRUs) and residual network (ResNet). Our model was evaluated on milk quality prediction dataset with seven unique feature such as pH, temperature, taste, odor, fat, turbidity, and color. The prediction output is classified with high (Goog), Low (Bad), and Medium (Moderate) classes. Our model shows superior results with comparison with multi-layer perceptron (MLP), random forest (RF) and support vector machine (SVM). In terms of accuracy, precision, recall, and F1-score, 0.996, 0.992, 0.992, 0.992.

This research explores the strategic optimization of secondary chlorination in water distribution systems (WDSs), in order to enhance the efficiency of disinfection while mitigating odor and operational costs and promoting sustainability in water quality management. The methodology integrates EPANET simulations for water hydraulic and quality modeling with a deep belief network (DBN) within the deep learning framework for accurate chloric odor prediction. Utilizing the non-dominated sorting genetic algorithm-II (NSGA-II), this methodology systematically balances the objectives of chloride dosage and chloramine formation. It combines a chloric odor intensity assessment, a multi-component kinetic model, and dual-objective optimization to conduct a comparative analysis of case studies on secondary chlorination strategies. The optimal configuration with five secondary chlorination stations reduced chloric odor intensity to 1.20 at a cost of USD 40,020.77 per year in Network A while, with eight stations, chloric odor intensity was reduced to 0.88 at a cost of USD 71,405.38 per year in Network B. The results demonstrate a balanced trade-off between odor intensity and operational cost on one hand and sustainability on the other hand, highlighting the importance of precise chlorine management to improve both the sensory and safety qualities of drinking water while ensuring the sustainable use and management of water resources.

Understanding the relationship between the chemical structure and physicochemical properties of odor molecules and olfactory perception, i.e., the structure–odor relationship, remains a decades-old, challenging task. However, the differences among the molecular structure graphs of different molecules are subtle and complex, and the molecular feature descriptors are numerous, with complex interactions that cause multiple odor perceptions. In this paper, we propose to decompose the features of the molecular structure graph into feature vectors corresponding to each odor perception descriptor to effectively explore higher-order semantic interactions between odor molecules and odor perception descriptors. We propose an olfactory perception prediction model noted as HGAFMN, which utilizes a hypergraph neural network with the olfactory lateral inhibition-inspired attention mechanism to learn the molecular structure feature from the odor molecular structure graph. Furthermore, existing methods cannot effectively extract interactive features in the large number of molecular feature descriptors, which have complex relations. To solve this problem, we add an attentional factorization mechanism to the deep neural network module and obtain a molecular descriptive feature through the deep feature combination based on the attention mechanism. Our proposed HGAFMN has achieved good results in extensive experiments and will help product design and quality assessment in the food, beverage, and fragrance industries.

One of the most important aspects of coffee's sensory experience is its scent, which is affected by roasting, microbial contamination, and place of origin. Electronic noses (ENs), which provide quick and precise identification of intricate odor patterns linked to microbial activity, have become highly effective instruments for evaluating coffee odors. This study examines the most recent uses of EN technology for evaluating coffee scent and emphasizes how important it is for assuring the safety and quality of the final product. The integration of ENs with artificial neural networks (ANNs) improves their precision in fragrance profile prediction and classification. Moreover, real-time smell profile monitoring is made possible by automated systems that use ENs, which helps with quality control in the coffee manufacturing process. Beyond the coffee business, EN technology is used in many other areas, including food safety, medical diagnostics, and environmental monitoring. Through innovation in fragrance analysis and quality control, this multidisciplinary approach promotes customer confidence and product integrity across industries.

<abstract><p>Odor is central to food quality. Still, a major challenge is to understand how the odorants present in a given food contribute to its specific odor profile, and how to predict this olfactory outcome from the chemical composition. In this proof-of-concept study, we seek to develop an integrative model that combines expert knowledge, fuzzy logic, and machine learning to predict the quantitative odor description of complex mixtures of odorants. The model output is the intensity of relevant odor sensory attributes calculated on the basis of the content in odor-active comounds. The core of the model is the mathematically formalized knowledge of four senior flavorists, which provided a set of optimized rules describing the sensory-relevant combinations of odor qualities the experts have in mind to elaborate the target odor sensory attributes. The model first queries analytical and sensory databases in order to standardize, homogenize, and quantitatively code the odor descriptors of the odorants. Then the standardized odor descriptors are translated into a limited number of odor qualities used by the experts thanks to an ontology. A third step consists of aggregating all the information in terms of odor qualities across all the odorants found in a given product. The final step is a set of knowledge-based fuzzy membership functions representing the flavorist expertise and ensuring the prediction of the intensity of the target odor sensory descriptors on the basis of the products' aggregated odor qualities; several methods of optimization of the fuzzy membership functions have been tested. Finally, the model was applied to predict the odor profile of 16 red wines from two grape varieties for which the content in odorants was available. The results showed that the model can predict the perceptual outcome of food odor with a certain level of accuracy, and may also provide insights into combinations of odorants not mentioned by the experts.</p></abstract>

Matsutake mushrooms, known for their high value, present challenges due to their seasonal availability, difficulties in harvesting, and short shelf life, making it crucial to extend their post-harvest preservation period. In this study, we developed three quality predictive models of Matsutake mushrooms using three different methods. The quality changes of Matsutake mushrooms were experimentally analyzed under two cases (case A: Temperature control and sealing measures; case B: Alteration of gas composition) with various parameters including the hardness, color, odor, pH, soluble solids content (SSC), and moisture content (MC) collected as indicators of quality changes throughout the storage period. Prediction models for Matsutake mushroom quality were developed using three different methods based on the collected data: multiple linear regression (MLR), support vector regression (SVR), and an artificial neural network (ANN). The comparative results reveal that the ANN outperforms MLR and SVR as the optimal model for predicting Matsutake mushroom quality indicators. To further enhance the ANN model’s performance, optimization techniques such as the Levenberg–Marquardt, Bayesian regularization, and scaled conjugate gradient backpropagation algorithm techniques were employed. The optimized ANN model achieved impressive results, with an R-Square value of 0.988 and an MSE of 0.099 under case A, and an R-Square of 0.981 and an MSE of 0.164 under case B. These findings provide valuable insights for the development of new preservation methods, contributing to the assurance of a high-quality supply of Matsutake mushrooms in the market.

Les mammifères identifient et interprètent une myriade de stimuli olfactifs par un mécanisme de codage complexe reposant sur la reconnaissance des molécules odorantes par des centaines de récepteurs olfactifs (RO). Ces interactions génèrent des combinaisons uniques de récepteurs activés, appelées code combinatoire, que le cerveau humain interprète comme la sensation que nous appelons l'odeur. Jusqu'à présent, le grand nombre de combinaisons possibles entre les récepteurs et les molécules a empêché une étude expérimentale à grande échelle de ce code et de son lien avec la perception des odeurs. La révélation de ce code est donc cruciale pour répondre à la question à long terme de savoir comment nous percevons notre environnement chimique complexe. Les RO appartiennent à la classe A des récepteurs couplés aux protéines G (RCPG) et constituent la plus grande famille multigénique connue. Pour étudier de façon systématique le codage olfactif, nous avons développé M2OR, une base de données exhaustive compilant les 25 dernières années d'essais biologiques sur les RO. À l'aide de cet ensemble de données, un modèle d'apprentissage profond sur mesure a été conçu et entraîné. Il combine l'intégration de jetons [CLS] d'un modèle de langage des protéines avec des réseaux de neurones en graphes et un mécanisme d'attention multi-têtes. Ce modèle prédit l'activation des RO par les odorants et révèle le code combinatoire résultant pour toute molécule odorante. Cette approche est affinée en développant un nouveau modèle capable de prédire l'activité d'un odorant à une concentration spécifique, permettant alors d'estimer la valeur d'EC50 de n'importe quelle paire OR-odorant. Enfin, les codes combinatoires dérivés des deux modèles sont utilisés pour prédire la perception olfactive des molécules. En incorporant des biais inductifs inspirés par la théorie du codage olfactif, un modèle d'apprentissage automatique basé sur ces codes est plus performant que l'état de l'art actuel en matière de prédiction d'odeurs. À notre connaissance, il s'agit de l'application la plus aboutie liant le code combinatoire à la prédiction de l'odeur d'une molécule. Dans l'ensemble, ce travail établit un lien entre les interactions complexes molécule odorante-récepteur et la perception humaine
Mammals identify and interpret a myriad of olfactory stimuli using a complex coding mechanism involving interactions between odorant molecules and hundreds of olfactory receptors (ORs). These interactions generate unique combinations of activated receptors, called the combinatorial code, which the human brain interprets as the sensation we call smell. Until now, the vast number of possible receptor-molecule combinations have prevented a large-scale experimental study of this code and its link to odor perception. Therefore, revealing this code is crucial to answering the long-term question of how we perceive our intricate chemical environment. ORs belong to the class A of G protein-coupled receptors (GPCRs) and constitute the largest known multigene family. To systematically study olfactory coding, we develop M2OR, a comprehensive database compiling the last 25 years of OR bioassays. Using this dataset, a tailored deep learning model is designed and trained. It combines the [CLS] token embedding from a protein language model with graph neural networks and multi-head attention. This model predicts the activation of ORs by odorants and reveals the resulting combinatorial code for any odorous molecule. This approach is refined by developing a novel model capable of predicting the activity of an odorant at a specific concentration, subsequently allowing the estimation of the EC50 value for any OR-odorant pair. Finally, the combinatorial codes derived from both models are used to predict the odor perception of molecules. By incorporating inductive biases inspired by olfactory coding theory, a machine learning model based on these codes outperforms the current state-of-the-art in smell prediction. To the best of our knowledge, this is the most comprehensive and successful application of combinatorial coding to odor quality prediction. Overall, this work provides a link between the complex molecule-receptor interactions and human perception

Over the past few decades, there has been a tremendous development in machine learning (ML), particularly in the areas of deep learning (DL) and transfer learning (TL). Deep learning has emerged as a powerful approach for solving complex problems in various domains such as computer vision, natural language processing, and speech recognition. At the same time, transfer learning has proven to be an effective technique for leveraging pre-trained deep learning models in new application domains with limited data. Design patterns, that are formalized best practices, offer a way to capture common problems and provide reusable solutions using generic and well-proven machine learning designs. This chapter aims to provide an overview of the advancements in deep learning and transfer learning, while emphasizing the significance of design patterns in addressing common challenges during the design of machine learning applications and systems. This work explores the implementation and results of various machine learning models on the mushroom classification dataset. The dataset comprises descriptions of 23 species of gilled mushrooms, with diverse features like cap shape, color, odor, and more. The goal was to classify mushrooms as edible, poisonous, or of unknown edibility. Among the models considered, the multi-layer perceptron (MLP), recurrent neural network (RNN), long short-term memory (LSTM), autoencoders, and Boltzmann machine were trained and evaluated. The MLP, RNN, and LSTM exhibited exceptional performance, achieving perfect training and testing accuracies of 1.0000. These models successfully learned the underlying patterns and features, resulting in accurate predictions on both training and shown test data. Deep learning can optimize mushroom waste processing by classifying waste types, optimizing composting conditions, and extracting nutrients for reuse, enhancing sustainability and resource recovery in agriculture. It also predicts market demand, automates quality control, and facilitates predictive maintenance, improving efficiency and reducing environmental impact.

The ozone process, which is the latter process of the water purification process, injects ozone to remove taste odor substances from tap water. Still, it is difficult to work the ozone process due to recent changes in water quality, such as taste and odor substances due to climate change. Therefore, this study developed an ozone injection rate determination model and a residual ozone concentration prediction model to properly remove flavor odor substances from raw water and proposed an operational diagnosis and optimal decision-making method for the ozone process in water purification. An ozone injection rate determination model and a residual ozone concentration prediction model were developed using data on water quality, flow rate, and operating conditions measured at Seoul's Y water purification plant. Two models were developed: the random forest and the MLP models. The performance difference between the two was verified by comparing the correlation coefficient and error index. Bayesian optimization, a global search method within a given composition space, was used to determine hyperparameters for each model. RMSE was selected as an objective function to determine the optimal hyperparameter through cross-validation. If the above model is applied to the ozone process, it is expected that an immediate response to changes in raw water quality and human error prevention will be possible.