Activity-specific Models Research Articles

Gait Phase Detection in Walking and Stairs Using Machine Learning.

Machine learning-based activity and gait phase recognition algorithms are used in powered motion assistive devices to inform control of motorized components. The objective of this study was to develop a supervised multiclass classifier to simultaneously detect activity and gait phase (stance, swing) in real-world walking, stair ascent, and stair descent using inertial measurement data from the thigh and shank. The intended use of this algorithm was for control of a motion assistive device local to the knee. Using data from 80 participants, two decision trees and five long short-term memory (LSTM) models that each used different feature sets were initially tested and evaluated using a novel performance metric: proportion of perfectly classified strides (PPCS). Based on the PPCS of these initial models, five additional posthoc LSTM models were tested. Separate models were developed to classify (i) both activity and gait phase simultaneously (one model predicting six states), and (ii) activity-specific models (three individual binary classifiers predicting stance/swing phases). The superior activity-specific model had an accuracy of 98.0% and PPCS of 55.7%. The superior six-phase model used filtered inertial measurement data as its features and a median filter on its predictions and had an accuracy of 92.1% and PPCS of 22.9%. Pooling stance and swing phases from all activities and treating this model as a binary classifier, this model had an accuracy of 97.1%, which may be acceptable for real-world lower limb exoskeleton control if only stance and swing gait phases must be detected. Keywords: machine learning, deep learning, inertial measurement unit, activity recognition, gait.

A TPACK-in Practice Model for Enhancing EFL Students’ Readiness to Teach with Ed-Tech Apps

<p class="0abstract">In Indonesia, educational technology courses have been offered in teacher preparation programs to equip students with theoretical knowledge and practical skill of utilizing technology in skill-based language teaching. Despite this, empirical studies focused on online training for EFL teacher candidates on using digital tools appropriately for pedagogical purposes and adjusting to traditional learning processes remain unexplored. Therefore, to address this gap, the present study attempts to investigate the design and implementation of a TPACK-in practice model to enhance students’ preparedness to face the rapidly-growing digital world and, more importantly, understand post-pandemic pedagogy. For this purpose, a qualitative case study was adopted. The data were obtained from multiple sources, including field notes, video-recorded classroom observation, one-on-one online interviews, and task documentation. The population of this research included 163 students in the 5<sup>th</sup> semester of the English Language Education Study Program of a university in <em>Yogyakarta</em>, Indonesia. We selected one class randomly, consisting of 40 pre-service English teachers during the odd semester of the 2020/2021 Academic Year. Findings demonstrate: 1) lesson activity-specific model in the educational technology course, 2) students’ reports on their hands-on learning experiences in designing teaching-learning activities with a variety of digital tools, 3) the perceived benefits of reviewing ed-tech apps, and 4) challenges encountered by students in completing their project. Finally, pedagogical implications are presented with proposed suggestions offered for future research directions in replicating the study.</p>

Prior automatic posture and activity identification improves physical activity energy expenditure prediction from hip-worn triaxial accelerometry.

Accelerometry is increasingly used to quantify physical activity (PA) and related energy expenditure (EE). Linear regression models designed to derive PAEE from accelerometry-counts have shown their limits, mostly due to the lack of consideration of the nature of activities performed. Here we tested whether a model coupling an automatic activity/posture recognition (AAR) algorithm with an activity-specific count-based model, developed in 61 subjects in laboratory conditions, improved PAEE and total EE (TEE) predictions from a hip-worn triaxial-accelerometer (ActigraphGT3X+) in free-living conditions. Data from two independent subject groups of varying body mass index and age were considered: 20 subjects engaged in a 3-h urban-circuit, with activity-by-activity reference PAEE from combined heart-rate and accelerometry monitoring (Actiheart); and 56 subjects involved in a 14-day trial, with PAEE and TEE measured using the doubly-labeled water method. PAEE was estimated from accelerometry using the activity-specific model coupled to the AAR algorithm (AAR model), a simple linear model (SLM), and equations provided by the companion-software of used activity-devices (Freedson and Actiheart models). AAR-model predictions were in closer agreement with selected references than those from other count-based models, both for PAEE during the urban-circuit (RMSE = 6.19 vs 7.90 for SLM and 9.62 kJ/min for Freedson) and for EE over the 14-day trial, reaching Actiheart performances in the latter (PAEE: RMSE = 0.93 vs. 1.53 for SLM, 1.43 for Freedson, 0.91 MJ/day for Actiheart; TEE: RMSE = 1.05 vs. 1.57 for SLM, 1.70 for Freedson, 0.95 MJ/day for Actiheart). Overall, the AAR model resulted in a 43% increase of daily PAEE variance explained by accelerometry predictions. NEW & NOTEWORTHY Although triaxial accelerometry is widely used in free-living conditions to assess the impact of physical activity energy expenditure (PAEE) on health, its precision and accuracy are often debated. Here we developed and validated an activity-specific model which, coupled with an automatic activity-recognition algorithm, improved the variance explained by the predictions from accelerometry counts by 43% of daily PAEE compared with models relying on a simple relationship between accelerometry counts and EE.

The Role of Heart-Rate Variability Parameters in Activity Recognition and Energy-Expenditure Estimation Using Wearable Sensors.

Human-activity recognition (HAR) and energy-expenditure (EE) estimation are major functions in the mobile healthcare system. Both functions have been investigated for a long time; however, several challenges remain unsolved, such as the confusion between activities and the recognition of energy-consuming activities involving little or no movement. To solve these problems, we propose a novel approach using an accelerometer and electrocardiogram (ECG). First, we collected a database of six activities (sitting, standing, walking, ascending, resting and running) of 13 voluntary participants. We compared the HAR performances of three models with respect to the input data type (with none, all, or some of the heart-rate variability (HRV) parameters). The best recognition performance was 96.35%, which was obtained with some selected HRV parameters. EE was also estimated for different choices of the input data type (with or without HRV parameters) and the model type (single and activity-specific). The best estimation performance was found in the case of the activity-specific model with HRV parameters. Our findings indicate that the use of human physiological data, obtained by wearable sensors, has a significant impact on both HAR and EE estimation, which are crucial functions in the mobile healthcare system.

Extending Participatory Sensing to Personal Exposure Using Microscopic Land Use Regression Models.

Personal exposure is sensitive to the personal features and behavior of the individual, and including interpersonal variability will improve the health and quality of life evaluations. Participatory sensing assesses the spatial and temporal variability of environmental indicators and is used to quantify this interpersonal variability. Transferring the participatory sensing information to a specific study population is a basic requirement for epidemiological studies in the near future. We propose a methodology to reduce the void between participatory sensing and health research. Instantaneous microscopic land-use regression modeling (µLUR) is an innovative approach. Data science techniques extract the activity-specific and route-sensitive spatiotemporal variability from the data. A data workflow to prepare and apply µLUR models to any mobile population is presented. The µLUR technique and data workflow are illustrated with models for exposure to traffic related Black Carbon. The example µLURs are available for three micro-environments; bicycle, in-vehicle, and indoor. Instantaneous noise assessments supply instantaneous traffic information to the µLURs. The activity specific models are combined into an instantaneous personal exposure model for Black Carbon. An independent external validation reached a correlation of 0.65. The µLURs can be applied to simulated behavioral patterns of individuals in epidemiological cohorts for advanced health and policy research.

Evaluation of custom energy expenditure models for SenseWear armband in manual wheelchair users.

Physical activity monitors are increasingly used to help the general population lead a healthy lifestyle by keeping track of their daily physical activity (PA) and energy expenditure (EE). However, none of the commercially available activity monitors can accurately estimate PA and EE in people who use wheelchairs as their primary means of mobility. Researchers have recently developed custom EE prediction models for manual wheelchair users (MWUs) with spinal cord injuries (SCIs) based on a commercial activity monitor--the SenseWear armband. This study evaluated the performance of two custom EE prediction models, including a general model and a set of activity-specific models among 45 MWUs with SCI. The estimated EE was obtained by using the two custom models and the default manufacturer's model, and it was compared with the gold standard measured by the K4b2 portable metabolic cart. The general, activity-specific, and default models had a mean signed percent error (mean +/- standard deviation) of -2.8 +/- 26.1%, -4.8 +/- 25.4%, and -39.6 +/- 37.8%, respectively. The intraclass correlation coefficient was 0.86 (95% confidence interval [CI] = 0.82 to 0.89) for the general model, 0.83 (95% CI = 0.79 to 0.87) for the activity-specific model, and 0.62 (95% CI = 0.16 to 0.81) for the default model. The custom models for the SenseWear armband significantly improved the EE estimation accuracy for MWUs with SCI.

Bankroto diagnostikos modelis ir jo pritaikymas bankroto tikimybei Lietuvos įmonėse prognozuoti

Businesses are constantly accompanied by uncertainty and risky decisions in dynamic and competitive market economy. In this context, corporate bankruptcies have become an unavoidable phenomenon the consequence of which is not only a problem of the companies themselves, but also determine the overall development of the country“™s economy. This process plays an important role in the economy because challenges companies to look for new ways to improve their performance and naturally remove inefficient participants from the market, promote competition. Various scientists have suggested many different bankrupt likelihood prediction models, but researches confirm that they are not entirely suitable for Lithuanian companies. In order to ensure business continuity it is important to prepare a model which could precisely predict bankrupt likelihood in Lithuanian companies. Both Lithuanian and foreign scientific studies focus on new specific bankrupt likelihood prediction models, but their application needs specific information. Scientific research results shows that most suitable models to predict bankrupt likelihood in Lithuanian companies are based on logistics and the multiple logistic regression. In this context, the new bankrupt likelihood prediction model is formed on the basis of logistic regression. Scientists developed various bankrupt likelihood prediction models at different times and in countries which differ in terms of development level, competitive conditions, and other characteristics. They are created using different financial data and different indicators. Country and activity specific models reduce the precision level at which they predict bankrupt likelihood in Lithuanian companies, so there is a practical problem ““ these models are not enough accurate for prediction. Solution to this problem, and thus the purpose of the research ““ to develop a model which allows accurately predict bankrupt likelihood in Lithuanian companies. The article is divided into three main parts: firstly we present analysis of bankruptcy diagnostic system; secondly, on the basis of theoretical analysis, we motivate logistic regression technique which will be employed to create bankrupt likelihood model for Lithuanian companies; thirdly, on the basis of empirical research, we present bankrupt likelihood model for Lithuanian companies also its activity and size specific variations and perform model quality and reliability assessment. In order to identify critical situations in Lithuanian companies at the earliest possible stage, it is recommended to apply in practice prepared bankrupt likelihood prediction model and its variations for companies operating in construction, industry, trade, service and other branch specific business and also models for size specific micro-small and medium-sized companies. Each of these models compatibility with the data and their quality are evaluated with statistical tests. The pattern of variation in predicting bankrupt likelihood of Lithuanian companies suggests using several models at once (industry specific and company size specific). For Lithuanian companies, depending on their branch or the size, to predict bankrupt likelihood, we can give such Z functions: Z construction branch = -1,094 ““ 5,330 (GAK/T) ““ 0,769 (PP/T) ““ 1,263 (TT/Tc®) + 9,059 (NK/c®) Z trade branch = -0,938 ““ 13,047 (GAK/T) + 0,064 (PP/GAK) ““ 2,368 (PP/T) + 12,772 (NK/c®) Z industry branch = -3,062 ““ 3,448 (GAK/T) ““ 1,234 (TT/Tc®) + 8,954 (NK/c®) Z service and other branches = -0,075 ““ 2,023 (GAK/T) ““ 2,176 (TT/Tc®) + 11,334 (NK/c®) Z micro-small sized = -2,191 ““ 2,504 (GAK/T) ““ 0,861 (TT/Tc®) + 6,425 (NK/c®) Z average sized = -4,025 ““ 8,956 (GAK/T) ““ 0,984 (PP/T) + 12,331 (NK/c®) ­­­­ These Z functions include financial coefficients, which can be calculated on the bases of information received from companies“™ financial data. According to the received Z value, bankrupt likelihood is calculated by the following formula: ; PN” [0;1] The resulting estimate of P indicates the probability for the company not to go bankrupt. According to the analyzed bankrupt likelihood prediction models based on logistic regression and received bankruptcy probability estimates form analyzed companies, we can provide such criteria for interpretation of analysis results: If P 0.75, the company will not go bankrupt. A composed model and its variations provide more accurate bankrupt likelihood prediction for Lithuanian companies than classic ones.

Predicting Energy Expenditure of Manual Wheelchair Users With Spinal Cord Injury Using a Multisensor-Based Activity Monitor

Hiremath SV, Ding D, Farringdon J, Cooper RA. Predicting energy expenditure of manual wheelchair users with spinal cord injury using a multisensor-based activity monitor. ObjectiveTo develop and evaluate new energy expenditure (EE) prediction models for manual wheelchair users (MWUs) with spinal cord injury (SCI) based on a commercially available multisensor-based activity monitor. DesignCross-sectional. SettingLaboratory. ParticipantsVolunteer sample of MWUs with SCI (N=45). InterventionSubjects were asked to perform 4 activities including resting, wheelchair propulsion, arm-ergometer exercise, and deskwork. Criterion EE using a metabolic cart and raw sensor data from a multisensor activity monitor was collected during each of these activities. Main Outcome MeasuresTwo new EE prediction models including a general model and an activity-specific model were developed using enhanced all-possible regressions on 36 MWUs and tested on the remaining 9 MWUs. ResultsThe activity-specific and general EE prediction models estimated the EE significantly better than the manufacturer's model. The average EE estimation error using the manufacturer's model and the new general and activity-specific models for all activities combined was –55.31% (overestimation), 2.30% (underestimation), and 4.85%, respectively. The average EE estimation error using the manufacturer's model, the new general model, and activity-specific models for various activities varied from –19.10% to –89.85%, –18.13% to 25.13%, and –4.31% to 9.93%, respectively. ConclusionsThe predictors for the new models were based on accelerometer and demographic variables, indicating that movement and subject parameters were necessary in estimating the EE. The results indicate that the multisensor activity monitor with new prediction models can be used to estimate EE in MWUs with SCI during wheelchair-related activities mentioned in this study.