Santoso Handri

Evaluation of Student's Physiological Response Towards E-Learning Courses Material by Using GSR Sensor

This study aims to evaluate student physiological response towards the e-learning materials. The experiments were conducted by introducing two contracting e-learning materials, i.e., the one is characterized as interactive material and the other is non-interactive one. During the experiment physiological sensor, i.e., galvanic skin response (GSR) sensor was attached to the participant. Furthermore, GSR data were extracted by feature generator, LDA. The purpose of feature extraction is to find preferably small number of features that are particularly distinguishing or informative for the classification process and that are invariant to irrelevant transformations of the data. Finally, several classifiers were performed discriminating student attitude towards e-learning course materials response using GSR sensor data. The results showed that discriminant analysis (DA) and support vector machine (SVM) give high accuracy rate, while the k-nearest neighbor (KNN) give moderate accuracy rate.

Determination of Age and Gender Based on Features of Human Motion Using AdaBoost Algorithms

Automated human identification by their walking behavior is a challenge attracting much interest among machine vision researchers. However, practical systems for such identification remain to be developed. In this study, a machine learning approach to understand human behavior based on motion imagery was proposed as the basis for developing pedestrian safety information systems. At the front end, image and video processing was performed to separate foreground from background images. Shape-width was then analyzed using 2D discrete wavelet transformation and 2D fast Fourier transformation to extract human motion features. Finally, an adaptive boosting (AdaBoost) algorithm was performed to classify human gender and age into its class based on spatiotemporal information. The results demonstrated the capability of the proposed systems to classify gender and age highly accurately.

Data mining the relationship between psychological and physiological stress measurement using rough set analysis

Recent behavioral medicine studies have revealed that some hormones secreted in human body sensitively changes according to his/her mental stress. Thus it is expected as an objective measurement of mental stress. However this newly developed interdisciplinary studies frequently showed inconsistent results. Some technical reasons were indicated for this discrepancy. Above all, we focused on the fact that in all these studies the only method introduced to estimate the relationship between the level of hormones and the subjective stress scale was the correlation analysis. In this study we employed Rough set analysis in place of conventional linear correlation analysis for mining the relationship between a subjective stress scale, “Profile of Mood Scale” (POMS), and a well known stress biomarker, salivary cortisol. As a developing result, numbers of items (inquiries in POMS) which relatively associated with cortisol level were found, whereas no significant linear correlation was obtained between them.

Development of a bionanodevice for detecting stress levels

Recent advances in molecular analysis techniques have enabled scientists to assess the tiny amounts of biochemical substances secreted in our bodies. This has revealed that the levels of various secretory hormones and immune substances vary sensitively with the mental state of a person. Such hormones and immune substances exhibit transient increases with various psychological stressors. They thus have the potential to be used as a novel biometric for monitoring stress. Biomarkers that occur in saliva can be monitored non-invasively and are thus potentially useful as practical indicators of mental stress. Stress biomarkers are considered to be released into the blood stream or other secretory fluids by physiological stress reactions. Stress biomarkers are expected to be detectable in sweat and other humoral fluids that are exuded from the skin surface. Based on this, we have developed a bionanodevice for detecting stress by capturing stress biomarkers on the skin surface in a non-invasive manner. A prototype bionanodevice is described in which a motor protein is introduced for molecular handling.

Evaluating Mental State of Final Year Students Based on POMS Questionnaire and HRV Signal

Final year students are normally encountering high pressing in their study. In view of this fact, this research focuses on determining mental states condition of college student in final year based on the psycho-physiological information. The experiments were conducted in two times, i.e., prior- and post- graduation seminar examination. The early results indicated that the student profile of mood states (POMS) in prior final graduation seminar showed higher scores than students in post final graduation seminar. Thus, in this research, relation between biosignal representing by heart rate variability (HRV) and questionnaire responses were evaluated by hidden Markov model (HMM) and neural networks (NN).

Evaluation of Accumulating Pattern of Human States Using HMM

The impact of extensively usage visual display terminal (VDT) on automatic nervous system are observed and analyzed based on physiological information, i.e., electroencephalogram (EEG), heart rate variability (HRV) and thermograph. Two types of experiment which have different intermittent schedule were conducted using the Kraepelin psycho diagnostic test. The purpose is to determine the given task towards effect of human mental states. The principal component analysis (PCA) and hidden Markov model (HMM) are then employed to differentiate the accumulation of human states condition among participants in two types of experiments. This experiment is conducted based on the assumption that the work schedule might have effect to the mental health problem.

Determination of Gender and Age Based on Pattern of Human Motion Using AdaBoost Algorithms

Automated human identification by their walking behavior is a challenge attracting much interest among machine vision researchers. However, practical systems for such identification remain to be developed. In this study, a machine learning approach to understand human behavior based on motion imagery was proposed as the basis for developing pedestrian safety information systems. At the front end, image and video processing was performed to separate foreground from background images. Shape-width was then analyzed using 2D discrete wavelet transformation to extract human motion features. Finally, an adaptive boosting (AdaBoost) algorithm was performed to classify human gender and age into its class. The results demonstrated capability of the proposed systems to classify gender and age highly accurately.

Adaptation of Learning Parameters for Choquet Integral Agent Network by Using Genetic Algorithms

Choquet Integral Agent Network (CHIAN) is proposed as a method realizing flexible information fusion which is constructed by using fuzzy measure and Choquet integrals. In case of multi-layered network structure, CHIAN can employ back-propagation algorithms-like concept for learning process. However, the back-propagation methods have some limitations such as trapping at local minima and network paralysis. Due to genetic algorithms (GA) mechanism, it has the characteristics of hill climbing, and thus can overcome the difficulty of trapping at local minima; consequently it might reduce network paralysis. This paper aims at proposing to tune CHIAN learning parameters, i.e., learning rate and momentum coefficient by genetic algorithms for improving CHIAN as classifier, pattern recognition, and information fusion. The results show that the network evolved GA requires fewer training cycles than the network which the learning parameters are intuitively given.