Development of an Artificial Neural Network Utilizing Particle Swarm Optimization for Modeling the Spray Drying of Coconut Milk.

Spray drying techniques are one of the methods to preserve and extend the shelf-life of coconut milk. The objective of this research was to create a particle swarm optimization-enhanced artificial neural network (PSO-ANN) that could predict the coconut milk spray drying process. The parameters for PSO tuning were selected as the number of particles and acceleration constant, respectively, for both global and personal best using a 2k factorial design. The optimal PSO settings were recorded as global best, C1 = 4.0; personal best, C2 = 0; and number of particles = 100. When comparing different types of spray drying models, PSO-ANN had an MSE value of 0.077, GA-ANN had an MSE of 0.033, while ANN had an MSE of 0.082. Sensitivity analysis was conducted on all three models to evaluate the significance level of each parameter on the model, and it was discovered that inlet temperature had the most significant influence on the model performance. In conclusion, the PSO-ANN was found to be more effective than ANN but less effective than GA-ANN in predicting the quality of coconut milk powder.

Features selection for Bayes classification of prolonged fatigue on rectus femoris muscle.

Sports training is very important to athletes in improving and maintaining their performances. It commonly involves high intensity exercise and requires longer time for fatigue to recover, compares to normal activity. During training, adequate rest is essential to allow recuperation and build body strength. Unfortunately, inadequate rest exposes the body to prolonged fatigue (PF). Hence, this condition needs to be managed accordingly to avoid chronic fatigue syndrome. Recent findings indicate that there are strong characteristics on surface EMG under PF conditions. Currently, the assessment is limited to glycogen breakdown, the existence of lactate and soreness. In this study, twenty participants were recruited to perform five days intensive training (IT) to induce more PF signs. The IT conducted was based on Bruce Protocol treadmill test. It was discovered that the IT successfully induces soreness, unexplained lethargy and performance decrement. Surface EMG were collected from rectus femoris muscle during daily pre and post treadmill tests. Four features were extracted from the surface EMG; mean frequency (Fmean), median frequency (Fmed), root mean square (RMS) and mean absolute value (MAV). The results indicated that all features during post exercise had greater value under PF condition and it was significant at P<;0.05. The features then were classified in accordance to Bayes. The results also showed that Fmed and MAV features offered good performance with 83.1% accuracy, 84.6% specificity and 80% of precision.

Amplitude and frequency changes in surface EMG of biceps femoris during five days Bruce Protocol treadmill test.

Electromyography (EMG) is one of the indirect tools in indexing fatigue. Fatigue can be detected when there are changes on amplitude and frequency. However, various outcomes from literature make researchers conclude that EMG is not a reliable tool to measure fatigue. This paper investigates EMG behavior of biceps femoris in median frequency and mean absolute value during five days of Bruce Protocol treadmill test. Before that, surface EMG signals are filtered using band pass filter cut-off at 20-500Hz and are de-noised using db45 1-decimated wavelet transform. Five participants achieved more than 85% of their maximal heart rate during the running activity. The authors also consider other markers of fatigue such as performance, muscle soreness and lethargy as indicators to adaptation and maladaptation conditions. Result shows that turning points of median frequency and mean absolute value are very significant in indexing fatigue and indicators to adaptation of resistive training.