Predicting the properties of wheat flour from grains during debranning: A machine learning approach.

In the food industry, meeting food quality demands is challenging. The quality of wheat flour, one of the most commonly used ingredients, depends on the extent of debranning done to remove the aleurone layer before milling. Therefore, the end product management can be simplified by predicting the properties of wheat flour during the debranning stage. Therefore, the chemical and rheological properties of grains were analyzed at different debranning durations (0, 30, 60 s). Then the images of wheat grain were taken to develop a regression model for predicting the chemical quality (i.e., ash, starch, fat, and protein contents) of the wheat flour. The resulting regression model comprises a convolutional neural network and is evaluated using the coefficient of determination (R 2), root-mean-square error, and mean absolute error as metrics. The results demonstrated that wheat flour contained more fat and protein and less ash with increasing debranning time. The model proved reliable in terms of root-mean-square error, mean absolute error, and R 2 for predicting ash content but not starch, fat, or protein contents, which can be attributed to the lack of features in the collected images of wheat kernels during debranning. In addition, the selected method, debranning, was beneficial to the rheological characteristics of wheat flour. The proportion of fine particles increased with the debranning time. The study experimentally revealed that the end product diversity for wheat flour can be controlled to provide selectable ingredients to customers.

A genetic programming-based optimal sensor placement for greenhouse monitoring and control.

Optimal sensor location methods are crucial to realize a sensor profile that achieves pre-defined performance criteria as well as minimum cost. In recent times, indoor cultivation systems have leveraged on optimal sensor location schemes for effective monitoring at minimum cost. Although the goal of monitoring in indoor cultivation system is to facilitate efficient control, most of the previously proposed methods are ill-posed as they do not approach optimal sensor location from a control perspective. Therefore in this work, a genetic programming-based optimal sensor placement for greenhouse monitoring and control is presented from a control perspective. Starting with a reference micro-climate condition (temperature and relative humidity) obtained by aggregating measurements from 56 dual sensors distributed within a greenhouse, we show that genetic programming can be used to select a minimum number of sensor locations as well as a symbolic representation of how to aggregate them to efficiently estimate the reference measurements from the 56 sensors. The results presented in terms of Pearson's correlation coefficient (r) and three error-related metrics demonstrate that the proposed model achieves an average r of 0.999 for both temperature and humidity and an average RMSE value of 0.0822 and 0.2534 for temperate and relative humidity respectively. Conclusively, the resulting models make use of only eight (8) sensors, indicating that only eight (8) are required to facilitate the efficient monitoring and control of the greenhouse facility.

Grid Search for Lowest Root Mean Squared Error in Predicting Optimal Sensor Location in Protected Cultivation Systems.

Irregular changes in the internal climates of protected cultivation systems can prevent attainment of optimal yield when the environmental conditions are not adequately monitored and controlled. Key to indoor environment monitoring and control and potentially reducing operational costs are the strategic placement of an optimal number of sensors using a robust method. A multi-objective approach based on supervised machine learning was used to determine the optimal number of sensors and installation positions in a protected cultivation system. Specifically, a gradient boosting algorithm, a form of a tree-based model, was fitted to measured (temperature and humidity) and derived conditions (dew point temperature, humidity ratio, enthalpy, and specific volume). Feature variables were forecasted in a time-series manner. Training and validation data were categorized without randomizing the observations to ensure the features remained time-dependent. Evaluations of the variations in the number and location of sensors by day, week, and month were done to observe the impact of environmental fluctuations on the optimal number and location of placement of sensors. Results showed that less than 32% of the 56 sensors considered in this study were needed to optimally monitor the protected cultivation system's internal environment with the highest occurring in May. In May, an average change of -0.041% in consecutive RMSE values ranged from the 1st sensor location (0.027°C) to the 17th sensor location (0.013°C). The derived properties better described the ambient condition of the indoor air than the directly measured, leading to a better performing machine learning model. A machine learning model was developed and proposed to determine the optimal sensors number and positions in a protected cultivation system.

Technological Advancements and Economics in Plant Production Systems: How to Retrofit?

Plant production systems such as plant factories and greenhouses can help promote resilience in food production. These systems could be used for plant protection and aid in controlling the micro- and macro- environments needed for optimal plant growth irrespective of natural disasters and changing climate conditions. However, to ensure optimal environmental controls and efficient production, several technologies such as sensors and robots have been developed and are at different stages of implementation. New and improved systems are continuously being investigated and developed with technological advances such as robotics, sensing, and artificial intelligence to mitigate hazards to humans working in these systems from poor ventilation and harsh weather while improving productivity. These technological advances necessitate frequent retrofits considering local contexts such as present and projected labor costs. The type of agricultural products also affects measures to be implemented to maximize returns on investment. Consequently, we formulated the retrofitting problem for plant production systems considering two objectives; minimizing the total cost for retrofitting and maximizing the yearly net profit. Additionally, we considered the following: (a) cost of new technologies; (b) present and projected cost for human labor and robotics; (c) size and service life of the plant production system; (d) productivity before and after retrofit, (e) interest on loans for retrofitting, (f) energy consumption before and after retrofit and, (g) replacement and maintenance cost of systems. We solved this problem using a multi-objective evolutionary algorithm that results in a set of compromised solutions and performed several simulations to demonstrate the applicability and robustness of the method. Results showed up to a 250% increase in annual net profits in an investigated case, indicating that the availability of all the possible retrofitting combinations would improve decision making. A user-friendly system was developed to provide all the feasible retrofitting combinations and total costs with the yearly return on investment in agricultural production systems in a single run.

Hybrid selection based multi/many-objective evolutionary algorithm.

In the last decade, numerous multi/many-objective evolutionary algorithms (MOEAs) have been proposed to handle multi/many-objective problems (MOPs) with challenges such as discontinuous Pareto Front (PF), degenerate PF, etc. MOEAs in the literature can be broadly divided into three categories based on the selection strategy employed such as dominance, decomposition, and indicator-based MOEAs. Each category of MOEAs have their advantages and disadvantages when solving MOPs with diverse characteristics. In this work, we propose a Hybrid Selection based MOEA, referred to as HS-MOEA, which is a simple yet effective hybridization of dominance, decomposition and indicator-based concepts. In other words, we propose a new environmental selection strategy where the Pareto-dominance, reference vectors and an indicator are combined to effectively balance the diversity and convergence properties of MOEA during the evolution. The superior performance of HS-MOEA compared to the state-of-the-art MOEAs is demonstrated through experimental simulations on DTLZ and WFG test suites with up to 10 objectives.

A Reference Vector-Based Simplified Covariance Matrix Adaptation Evolution Strategy for Constrained Global Optimization.

During the last two decades, the notion of multiobjective optimization (MOO) has been successfully adopted to solve the nonconvex constrained optimization problems (COPs) in their most general forms. However, such works mainly utilized the Pareto dominance-based MOO framework while the other successful MOO frameworks, such as the reference vector (RV) and the decomposition-based ones, have not drawn sufficient attention from the COP researchers. In this article, we utilize the concepts of the RV-based MOO to design a ranking strategy for the solutions of a COP. We first transform the COP into a biobjective optimization problem (BOP) and then solve it by using the covariance matrix adaptation evolution strategy (CMA-ES), which is arguably one of the most competitive evolutionary algorithms of current interest. We propose an RV-based ranking strategy to calculate the mean and update the covariance matrix in CMA-ES. Besides, the RV is explicitly tuned during the optimization process based on the characteristics of COPs in a RV-based MOO framework. We also propose a repair mechanism for the infeasible solutions and a restart strategy to facilitate the population to escape from the infeasible region. We test the proposal extensively on two well-known benchmark suites comprised of 36 and 112 test problems (at different scales) from the IEEE CEC (Congress on Evolutionary Computation) 2010 and 2017 competitions along with a real-world problem related to power flow. Our experimental results suggest that the proposed algorithm can meet or beat several other state-of-the-art constrained optimizers in terms of the performance on a wide variety of problems.

A Deep Learning Framework Integrating the Spectral and Spatial Features for Image-Assisted Medical Diagnostics.

The development of a computer-aided disease detection system to ease the long and arduous manual diagnostic process is an emerging research interest. Living through the recent outbreak of the COVID-19 virus, we propose a machine learning and computer vision algorithms-based automatic diagnostic solution for detecting the COVID-19 infection. Our proposed method applies to chest radiograph that uses readily available infrastructure. No studies in this direction have considered the spatial aspect of the medical images. This motivates us to investigate the role of spectral-domain information of medical images along with the spatial content towards improved disease detection ability. Successful integration of spatial and spectral features is demonstrated on the COVID-19 infection detection task. Our proposed method comprises three stages - Feature extraction, Dimensionality reduction via projection, and prediction. At first, images are transformed into spectral and spatio-spectral domains by using Discrete cosine transform (DCT) and Discrete Wavelet transform (DWT), two powerful image processing algorithms. Next, features from spatial, spectral, and spatio-spectral domains are projected into a lower dimension through the Convolutional Neural Network (CNN), and those three types of projected features are then fed to Multilayer Perceptron (MLP) for final prediction. The combination of the three types of features yielded superior performance than any of the features when used individually. This indicates the presence of complementary information in the spectral domain of the chest radiograph to characterize the considered medical condition. Moreover, saliency maps corresponding to classes representing different medical conditions demonstrate the reliability of the proposed method. The study is further extended to identify different medical conditions using diverse medical image datasets and shows the efficiency of leveraging the combined features. Altogether, the proposed method exhibits potential as a generalized and robust medical image-assisted diagnostic solution.

Spatial filter and feature selection optimization based on EA for multi-channel EEG.

The EEG signals employed for BCI systems are generally band-limited. The band-limited multiple Fourier linear combiner (BMFLC) with Kalman filter was developed to obtain amplitude estimates of the EEG signal in a pre-fixed frequency band in real-time. However, the high-dimensionality of the feature vector caused by the application of BMFLC to multi-channel EEG based BCI deteriorates the performance of the classifier. In this work, we apply evolutionary algorithm (EA) to tackle this problem. The real-valued EA encodes both the spatial filter and the feature selection into its solution and optimizes it with respect to the classification error. Three BMFLC based BCI configurations are proposed. Our results show that the BMFLC-KF with covariance matrix adaptation evolution strategy (CMAES) has the best overall performance.