A Learning Machine Approach for Predicting Thermal Comfort Indices

Numerical Investigation on the Relationship between Human Thermal Comfort and Thermal Balance under Radiant Cooling System

Chlorination has made water supply safe from bacteria and viruses, and has almost completely eliminated risks of waterborne diseases such as typhoid fever, cholera and dysentery. However, the health benefit of chlorination has introduced some possible risks from the by-products of the disinfection process. In this work, we propose the use of a learning algorithm ‘support vector machine (SVM)’ for the prediction of the water pollutant total exposure index (TEI) and cumulative risk (CR). The SVM is an artificial intelligent approach that could capture the input/output mapping from the given data. Support vector machines were developed based on the structural risk minimization principle from statistical learning theory in which the empirical risk and the machine complexity are optimized. The proposed algorithm was validated by using statistical measurements on experimental data different from those used for the training (out-of-sample testing) to evaluate the algorithm's ‘generalizability’. The experimental results demonstrate a good correlation between the SVM predicted values and those conventionally estimated using probabilistic (Monte Carlo) simulations. In both cases, the TEI and CR values were estimated using a set of data extracted from the national database IRIS. The SVM, once trained with representative data, could be used easily and effectively in comparison with other conventional estimation methods of TEI and CR.

A new prediction method for thermal comfort indices is introduced. This method of prediction titled ‘support vector machine (SVM)’ uses learning as a process to emulate human intelligence. In this paper, more adequate nonlinear kernels have been used and the SVM has been improved to predict the thermal comfort indices accurately. In this study, we focus mainly on supervised learning machine where an instructor provides the output samples during the learning phase. Different sets of representative experimental factors, such as air temperature, mean radiant temperature, relative humidity, air velocity, metabolism and clothing value that affect a person’s thermal balance were used for training the SVM machine. The results show the best correlation between SVM predicted values with a polynomial kernel of the second order and those obtained from conventional thermal comfort, such as the Fanger model and the ‘2-Node’ model.

Fuzzy comprehensive evaluation of human thermal comfort in simulating natural wind environment

The prediction model of indoor thermal comfort PMV index based on least squares support vector machine (LS-SVM) is established by using the nonlinear relationship between human thermal comfort and its influencing factors and the characteristic that particle swarm has of fast global optimization. Adopting the parameters of least squares support vector machine optimized by Particle Swarm algorithm, the mapping relations between the six factors including indoor air temperature, relative humidity, air velocity, mean radiant temperature, human metabolic rate, thermal resistance and PMV index can be formed through the sample data learning. The experimental results show that the method is accurate and effective.

The existing research on waterfront tour route planning suggests that thermal comfort assessment has a direct impact on people’s outdoor tour experiences. It contributes to understanding the thermal comfort regulation mechanisms of landscape nodes and the scientific planning of tour routes. However, the research in this area is relatively limited and faces complex issues in data processing and analysis. This study takes the West Lake waterfront tour route in Hangzhou, China, as an example. Through empirical research, it validates the feasibility of tour route planning guided by thermal comfort. It employs field surveys, on-site measurements, and data analysis methods primarily based on the Rayman software to explore the relationship between urban outdoor waterfront spaces and microclimates. The study reveals the following findings: (i) Extensive water bodies surrounding an area positively regulate local microclimates. The closer one is to the water and the more water surrounds the area, the stronger the microclimate effect. (ii) Factors such as spatial openness, waterfront orientation, prevailing monsoons, etc. all have certain impacts on the microclimates of the lake and lakeside spaces. (iii) Air temperature has significant variability in its impact on human thermal comfort, with different PET (Physiological Equivalent Temperature) values at each measurement point during different time periods. The study combines the actual human thermal comfort conditions at eight measurement points around the lake with existing tour routes. It proposes two practical tour route optimization designs based on real tour microclimate factors. These designs employ thermal comfort assessment methods to effectively enhance and improve human thermal comfort during tours. This research carries social value and provides new insights for mitigating urban waterfront space microclimates, enhancing tour comfort, and improving landscape environments.

To address the critical need for accurate human thermal comfort prediction in winter heating environments, this study established a comprehensive thermal comfort dataset containing 2089 valid samples through experiments. On this basis, thermal comfort prediction models were constructed using three multi-class machine learning algorithms: Support Vector Classification, K-Nearest Neighbors, and Random Forest. The predictive performance of 63 different feature combinations was systematically evaluated. The results indicate that the feature subset comprising indoor air temperature, forehead temperature, cheek temperature, dorsal hand temperature, heart rate, and systolic blood pressure yields the optimal prediction performance. Among the evaluated models, the Random Forest model demonstrated superior overall performance, achieving an accuracy exceeding 90% and an AUC ranging from 96% to 99%, significantly outperforming the SVC and KNN models. Compared with the traditional Predicted Mean Vote (PMV) model, the machine learning models developed in this study showed a substantial improvement in prediction accuracy under identical conditions; notably, the Random Forest model improved accuracy by approximately 40% over the PMV model. Based on these findings, a smart heating system framework integrating environmental sensors, wearable devices, and intelligent control valves is proposed, providing a theoretical basis and technical approach for realizing personalized and energy-efficient heating control.

Support vector machine (SVM) and artificial neural network (ANN) systems were applied to a drug/nondrug classification problem as an example of binary decision problems in early-phase virtual compound filtering and screening. The results indicate that solutions obtained by SVM training seem to be more robust with a smaller standard error compared to ANN training. Generally, the SVM classifier yielded slightly higher prediction accuracy than ANN, irrespective of the type of descriptors used for molecule encoding, the size of the training data sets, and the algorithm employed for neural network training. The performance was compared using various different descriptor sets and descriptor combinations based on the 120 standard Ghose-Crippen fragment descriptors, a wide range of 180 different properties and physicochemical descriptors from the Molecular Operating Environment (MOE) package, and 225 topological pharmacophore (CATS) descriptors. For the complete set of 525 descriptors cross-validated classification by SVM yielded 82% correct predictions (Matthews cc = 0.63), whereas ANN reached 80% correct predictions (Matthews cc = 0.58). Although SVM outperformed the ANN classifiers with regard to overall prediction accuracy, both methods were shown to complement each other, as the sets of true positives, false positives (overprediction), true negatives, and false negatives (underprediction) produced by the two classifiers were not identical. The theory of SVM and ANN training is briefly reviewed.

Lithology prediction by support vector classifiers using inverted seismic attributes data and petrophysical logs as a new approach and investigation of training data set size effect on its performance in a heterogeneous carbonate reservoir

Effect of thermal transient on human thermal comfort in temporarily occupied space in winter – A case study in Tianjin

The outdoor environment of subtropical islands with high temperature, high humidity, strong radiation and strong wind speed has a great impact on outdoor thermal safety and thermal comfort of personnel. Ensuring the thermal safety and thermal comfort of the human body is of great significance to the development of island tourism. This study conducted thermal environment monitoring and thermal comfort questionnaire experiments in Hailing Island Beach Scenic Area to explore the outdoor thermal environment and human thermal comfort in the beach area, identify the factors affecting human thermal comfort, and determine the thermal comfort benchmark for tourists in the beach area. The research results show that: 1) the beach underlying surface albedo is 0.4, which is much higher than the common urban underlying surface albedo. The beach radiation environment is strong and the thermal environment is poor in summer. 2) On sea-land breeze days, sea breeze can effectively alleviate the heat stress in the offshore areas of island beaches. 3) Forests have a significant attenuation effect on solar radiation. Even sparse coconut forests have a certain hindering effect on radiation and the thermal comfort of human beings in the forest space is relatively good. 4) Tourists’ active thermal adaptation behaviors can avoid thermal discomfort, but the effect is limited. 5) This study obtained the thermal comfort benchmarks for beach tourists, and the thermal neutral ranges were: 20.93-29.01 °C (PET), 25.91-30.50 °C (UTCI), and -29.13-50.24 W/m2 (COMFA), respectively. The research conclusions can provide a theoretical basis for the creation of outdoor thermal environment in beach scenic areas and the assessment of tourists' thermal comfort.

Bond rating is regarded as an important event for measuring financial risk of companies and for determining the investment returns of investors. As a result, it has been a popular research topic for researchers to predict companies' credit ratings by applying statistical and machine learning techniques. The statistical techniques, including multiple regression, multiple discriminant analysis (MDA), logistic models (LOGIT), and probit analysis, have been traditionally used in bond rating. However, one major drawback is that it should be based on strict assumptions. Such strict assumptions include linearity, normality, independence among predictor variables and pre-existing functional forms relating the criterion variablesand the predictor variables. Those strict assumptions of traditional statistics have limited their application to the real world. Machine learning techniques also used in bond rating prediction models include decision trees (DT), neural networks (NN), and Support Vector Machine (SVM). Especially, SVM is recognized as a new and promising classification and regression analysis method. SVM learns a separating hyperplane that can maximize the margin between two categories. SVM is simple enough to be analyzed mathematical, and leads to high performance in practical applications. SVM implements the structuralrisk minimization principle and searches to minimize an upper bound of the generalization error. In addition, the solution of SVM may be a global optimum and thus, overfitting is unlikely to occur with SVM. In addition, SVM does not require too many data sample for training since it builds prediction models by only using some representative sample near the boundaries called support vectors. A number of experimental researches have indicated that SVM has been successfully applied in a variety of pattern recognition fields. However, there are three major drawbacks that can be potential causes for degrading SVM's performance. First, SVM is originally proposed for solving binary-class classification problems. Methods for combining SVMs for multi-class classification such as One-Against-One, One-Against-All have been proposed, but they do not improve the performance in multi-class classification problem as much as SVM for binary-class classification. Second, approximation algorithms (e.g. decomposition methods, sequential minimal optimization algorithm) could be used for effective multi-class computation to reduce computation time, but it could deteriorate classification performance. Third, the difficulty in multi-class prediction problems is in data imbalance problem that can occur when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed boundary and thus the reduction in the classification accuracy of such a classifier. SVM ensemble learning is one of machine learning methods to cope with the above drawbacks. Ensemble learning is a method for improving the performance of classification and prediction algorithms. AdaBoost is one of the widely used ensemble learning techniques. It constructs a composite classifier by sequentially training classifiers while increasing weight on the misclassified observations through iterations. The observations that are incorrectly predicted by previous classifiers are chosen more often than examples that are correctly predicted. Thus Boosting attempts to produce new classifiers that are better able to predict examples for which the current ensemble's performance is poor. In this way, it can reinforce the training of the misclassified observations of the minority class. This paper proposes a multiclass Geometric Mean-based Boosting (MGM-Boost) to resolve multiclass prediction problem. Since MGM-Boost introduces the notion of geometric mean into AdaBoost, it can perform learning process considering the geometric mean-based accuracy and errors of multiclass. This study applies MGM-Boost to the real-world bond rating case for Korean companies to examine the feasibility of MGM-Boost. 10-fold cross validations for threetimes with different random seeds are performed in order to ensure that the comparison among three different classifiers does not happen by chance. For each of 10-fold cross validation, the entire data set is first partitioned into tenequal-sized sets, and then each set is in turn used as the test set while the classifier trains on the other nine sets. That is, cross-validated folds have been tested independently of each algorithm. Through these steps, we have obtained the results for classifiers on each of the 30 experiments. In the comparison of arithmetic mean-based prediction accuracy between individual classifiers, MGM-Boost (52.95%) shows higher prediction accuracy than both AdaBoost (51.69%) and SVM (49.47%). MGM-Boost (28.12%) also shows the higher prediction accuracy than AdaBoost (24.65%) and SVM (15.42%)in terms of geometric mean-based prediction accuracy. T-test is used to examine whether the performance of each classifiers for 30 folds is significantly different. The results indicate that performance of MGM-Boost is significantly different from AdaBoost and SVM classifiers at 1% level. These results mean that MGM-Boost can provide robust and stable solutions to multi-classproblems such as bond rating.

As an important evaluation index of indoor environmental quality, human thermal comfort is affected by various factors. As human physiological signal, the skin temperature is closely related to human thermal comfort. To investigate the model's mapping relationship between the skin temperature and the human thermal comfort, the skin temperatures of six parts of human body were chosen as research object, the linear mapping relationship between the average skin temperature and the human thermal comfort was analyzed by using Fisher discriminant method under five different experimental conditions and subjective voting. As a result, the mathematical evaluating model between the average skin temperature and human thermal comfort was obtained based on experiments. The results indicate the average precision of this model is high and the discriminant result is reliable, which provides valuable reference for human thermal comfort modeling under homogeneous environmental condition.

Since thermal comfort is becoming a frontier science combined with pyrology,ar-chitectural physics,physiology and psychology, as human important physiological indexes, skin temperature and heart rate variation are explored for the thermal comfort study in low-pressure environment. Important information of human thermal balance and thermal sensation can be obtained by testing the skin temperature, the change of heart rate (HR) can reflect human metabolic rate-an important influencing factor on human thermal comfort. Experiments were performed to investigate how these physiological factors respond to the environmental temperatures，especially pressure, and what is the relationship between skin temperature and heart rate variation (HRV) and thermal comfort. The respond of human skin temperature and HR to pressure and their influence on human mean thermal sensation (MTS) were studied by questionnaire and controlled experiments in the asymptomatic altitude reaction regions. The experimental results indicate that mean skin temperature and HR all decrease with the pressure drop when other environmental parameters kept constant, and MTS has the same tendency. It provides certain physiological basis for human thermal comfort study under low pressure.

Air-conditioning equipment is used in various places such as houses, office buildings, and public facilities and is indispensable in modern-day life. Therefore, the energy consumption of air-conditioning equipment accounts for a large percentage of the total energy consumption in the average household. Specifically, it accounts for 26% of the annual energy consumption in ordinary homes and 27% in industry, according to the Annual Energy Report for Japan, which was presented by the Ministry of the Economy, Trade, and Industry, and the Agency for Natural Resources and Energy in 2010. Therefore, it is desirable to reduce energy consumption by reducing the air-conditioning load. The Ministry of the Environment recommends a constant preset temperature of 28°C in summer to decrease energy consumption. However, many people feel uncomfortable in such a thermal environment. Thus, an air-conditioning control to simultaneously suppress energy consumption and maintain human thermal comfort is desired. To develop such a control, an index to accurately evaluate human thermal comfort is needed. When a person feels comfortable or uncomfortable, their prefrontal area, which is involved in thinking and the feeling of emotions, is activated. It is presumed that the measurement of the brain activation reaction of a person will reveal whether the person feels comfortable or uncomfortable in the thermal environment. The evaluation of thermal comfort by means of brain activation reactions will allow one to develop the optimum air-conditioning control to maintain human thermal comfort. This paper proposes a method to evaluate thermal comfort via brain signals and ultimately aims to develop an air-conditioning control system utilizing this evaluation method. This paper will describe the measurement procedure of brain activation reactions to indoor-temperature change by using near-infrared spectroscopy and the relationship between thermal comfort and brain activation reaction. This study also investigated the changes in oxyHb levels together with indoor-temperature changes, measured with the NIRS. We measured the changes in the oxyHb levels of the prefrontal area when the temperature increased and decreased. As a result, the oxyHb level in the prefrontal area correlated with the indoor-temperature change, the PMV, and the subjects’ declaration of thermal sensation. Conversely, the change in the oxyHb level with the inclusion of wind and a constant indoor temperature significantly differed with that with a varying indoor temperature. Furthermore, the oxyHb change correlated with the PMV and the subject’s declaration of thermal sensation. Therefore, the measured oxyHb change may represent the thermal comfort of a person.