Search

Facial expression recognition technology is significant to psychological analysis, polygraph technology, clinical medicine, and safe driving. However, traditional facial expression recognition technologies mainly recognize frontal facial expressions, and few studies focus on the recognition of facial expressions under deflection angles. In this situation, this paper combines Adaboost cascaded detectors, Ensemble of Regression Trees algorithm, and Random Forest algorithm to recognize facial expressions under deflection angles. After detecting faces by Adaboost cascaded detectors, the facial expression features under deflection angles are extracted by the Ensemble of Regression Trees algorithm. Finally, the Random Forest algorithm is used to output the recognition results based on expression features. Experiments show that the enhanced facial expression recognition algorithm proposed in this paper realizes the recognition of non-frontal facial expressions, and it has good real-time performance and recognition accuracy.

Existing diagnosis of the autism spectrum disorder (ASD) heavily depends on the informant’s evaluation of the patient’s behavior, which is both time consuming and labor demanding. In order to develop a rapid diagnostic tool with high accuracy, machine learning (ML) approaches have been proposed to explore the feasibility of identifying ASD with a limited number of features extracted from behavioral evaluation, neuroimaging and kinematic data. Though restricted and repetitive behavior (RRB) is one of the cardinal symptoms of ASD, no study has been conducted to investigate whether restricted kinematic features (RKF) could be used to identify ASD. The present study aimed to address this question. Twenty children with high functioning autism and twenty-three children with typical development (TD) were recruited. They were instructed to perform a motor task that required the execution of the utmost variant movement. Entropy and 95% range of the movement amplitude, velocity and acceleration were computed as indices of RKF. Five ML classifiers were trained and tested including support vector machine (SVM), Linear Discriminant Analysis (LDA), Decision tree (DT), Random forest (RF), and K nearest neighbor (KNN). Results showed that the KNN algorithm ( $k = 1$ ) yielded the highest classification accuracy with four kinematic features (accuracy: 88.37%, specificity: 91.3%, sensitivity: 85%, AUC: 0.8815). Our study demonstrated that RKF could help robustly identify ASD. It is inferred that the application of ML on genetic, neuroimaging, psychological and kinematic features might pose a considerable challenge to the current diagnostic criteria of ASD, and might potentially lead to an automated and objective diagnosis of ASD.

As the prodromal stage of Alzheimer's disease, effective recognition of mild cognitive impairment can reduce the prevalence of Alzheimer's disease. At present, most of the research on the recognition of mild cognitive impairment is carried out through biomarkers and neuroimaging, which is not conducive to large-scale analysis and research. Based on neuropsychological evaluation and life habits questionnaires, this article applies machine learning methods to the recognition of mild cognitive impairment and conducts experimental research. A questionnaire survey of the elderly was conducted to obtain raw data, including demographic variables, daily habits and neuropsychological data. Feature selection is carried out through filter screening method and the influence of factors such as lifestyle, physical health and learning ability on the morbidity of mild cognitive impairment is analyzed. The classifier mainly uses three methods: artificial neural network, support vector machine and random forest. The experimental results show that the random forest classification effect is the best and the accuracy rate is as high as 92%. Artificial neural network has strong generalization ability with 91% accuracy rate. Support vector machine has inferior effect.

Grid division can transform a complex physical model into several simple entities, and is an effective method for power distribution system planning and the whole process management. It also provides a new perspective for the optimal layout planning of switching stations as well as the development of computer-aided planning software packages. This paper first defines the demand grid for low-voltage electricity access demand. Then, an optimization model of switching station layout planning based on grid division is proposed, with the objectives of maximizing the average equipment utilization rate and minimizing the total number of switching stations, and with five constraints including the block boundary, crossing, power supply radius, planned capacity of the site, and the maximum load rate of the equipment. Besides, this paper further describes the algorithm flowchart developed based on this method and the application in the layout planning of actual medium-voltage switchyards.

Influenced by the ever-increasing electricity consumption and ambient temperature, the proportion of the air-conditioning load reaches up to 50% in China and shows a trend of further growth, which gives rise to the concerns of formulating effective measures to control the load growth and improve the load profile. The existing air-conditioning load forecasting methods are mostly for the total amount in a given area, rather than for a specific user or a block of users, and do not distinguish the users’ specific types, and hence lead to less accurate prediction results. Given this background, a load forecasting method for point loads is developed based on seasonal decomposition and the AutoRegressive Integrated Moving Average (ARIMA) model. Specifically, a correction method for seasonal component based on the maximum air-conditioning load is presented, and the complete annual air-conditioning load curve can then be obtained. In this way, the accuracy of load forecasting is effectively improved. This forecasting method can be widely used in applications such as the planning of distribution systems, microgrids, and virtual power plants, and can also provide an effective decision-making basis for formulating actual demand-side response strategies and developing control measures to regulate the air-conditioning load, and hence to reduce the peak load.