USING DATA SCIENCE TO EVALUATE NANO-REINFORCED EPOXY SURFACES

2021 ◽  
Author(s):  
JONATHAN THEIM ◽  
DANIEL P. COLE ◽  
UTKARSH DUBEY ◽  
ASHUTOSH SRIVASTAVA ◽  
CHOWDHURY ASHRAF ◽  
...  
Keyword(s):  
Data Science ◽  
Mechanical Performance ◽  
Narrow Range ◽  
Bayesian Optimization ◽  
Training Set ◽  
Full Field ◽  
Force Microscopy ◽  
Particle Position ◽  
Full Field Measurement ◽  
Using Data

Toughened composites reinforced with nanofillers show improved mechanical performance such as increased abrasion resistance, fracture toughness, and fracture energy. The degree of these improvements is influenced by the degree of dispersion of the nanofillers which can be analyzed using force microscopy (AFM), a technique that allows for mapping the local height and elastic modulus of a surface. However, current AFM apparatuses can only measure a narrow range of moduli according to the type of tip, which complicates the full-field measurement of moduli in nanocomposites with nanosilica (~72 GPa) embedded in epoxy (0.1 – 5 GPa). Moreover, height mapping can only visualize filler particles exposed at the surface. These limitations make it challenging to determine the 3D location of nanoparticles near the surface of a composite. To overcome these limitations of conventional AFM, we used a combination of data science, micromechanics, and experimental data from AFM to locate the centroidal position of nanosilica (NS) particles relative to the surrounding epoxy surface. Using finite element simulations, a theoretical dataset of modulus values as a function of particle position relative to the epoxy surface was created as a training set. Bayesian optimization determines the “best” particle position that results in minimum error between simulated and experimental modulus contours. The algorithm returns the 3D position of the fully or partially embedded NS particle relative to the epoxy surface. The algorithm has shown the ability to partially produce simulated modulus contours that resemble the experimental modulus contours.

scholarly journals UAV Image Multi-Labeling with Data-Efficient Transformers

2021 ◽  
Vol 11 (9) ◽  
pp. 3974
Author(s):  
Laila Bashmal ◽  
Yakoub Bazi ◽  
Mohamad Mahmoud Al Rahhal ◽  
Haikel Alhichri ◽  
Naif Al Ajlan
Keyword(s):  
Data Augmentation ◽  
Feature Representation ◽  
Aerial Image ◽  
Remote Sensing Images ◽  
Training Set ◽  
Proposed Model ◽  
Class Labels ◽  
Using Data ◽  
Uav Image

In this paper, we present an approach for the multi-label classification of remote sensing images based on data-efficient transformers. During the training phase, we generated a second view for each image from the training set using data augmentation. Then, both the image and its augmented version were reshaped into a sequence of flattened patches and then fed to the transformer encoder. The latter extracts a compact feature representation from each image with the help of a self-attention mechanism, which can handle the global dependencies between different regions of the high-resolution aerial image. On the top of the encoder, we mounted two classifiers, a token and a distiller classifier. During training, we minimized a global loss consisting of two terms, each corresponding to one of the two classifiers. In the test phase, we considered the average of the two classifiers as the final class labels. Experiments on two datasets acquired over the cities of Trento and Civezzano with a ground resolution of two-centimeter demonstrated the effectiveness of the proposed model.

scholarly journals New dilemmas, old problems: advances in data analysis and its geoethical implications in groundwater management

2021 ◽  
Vol 3 (6) ◽  
Author(s):  
César de Oliveira Ferreira Silva ◽  
Mariana Matulovic ◽  
Rodrigo Lilla Manzione
Keyword(s):  
Decision Making ◽  
Data Analytics ◽  
Data Science ◽  
Final Decision ◽  
Aquifer Recharge ◽  
Groundwater Governance ◽  
Ethical Attitude ◽  
New Information ◽  
Modeling Experiment ◽  
Using Data

Abstract Groundwater governance uses modeling to support decision making. Therefore, data science techniques are essential. Specific difficulties arise because variables must be used that cannot be directly measured, such as aquifer recharge and groundwater flow. However, such techniques involve dealing with (often not very explicitly stated) ethical questions. To support groundwater governance, these ethical questions cannot be solved straightforward. In this study, we propose an approach called “open-minded roadmap” to guide data analytics and modeling for groundwater governance decision making. To frame the ethical questions, we use the concept of geoethical thinking, a method to combine geoscience-expertise and societal responsibility of the geoscientist. We present a case study in groundwater monitoring modeling experiment using data analytics methods in southeast Brazil. A model based on fuzzy logic (with high expert intervention) and three data-driven models (with low expert intervention) are tested and evaluated for aquifer recharge in watersheds. The roadmap approach consists of three issues: (a) data acquisition, (b) modeling and (c) the open-minded (geo)ethical attitude. The level of expert intervention in the modeling stage and model validation are discussed. A search for gaps in the model use is made, anticipating issues through the development of application scenarios, to reach a final decision. When the model is validated in one watershed and then extrapolated to neighboring watersheds, we found large asymmetries in the recharge estimatives. Hence, we can show that more information (data, expertise etc.) is needed to improve the models’ predictability-skill. In the resulting iterative approach, new questions will arise (as new information comes available), and therefore, steady recourse to the open-minded roadmap is recommended. Graphic abstract

scholarly journals Raman and Infrared Thermometry for Microsystems

2013 ◽  
Vol 5 (3) ◽  
Author(s):  
Leslie M. Phinney ◽  
Wei-Yang Lu ◽  
Justin R. Serrano
Keyword(s):  
Data Collection ◽  
Spatial Resolution ◽  
Infrared Imaging ◽  
Silicon On Insulator ◽  
Pixel Size ◽  
Full Field ◽  
Infrared Thermometry ◽  
Full Field Measurement ◽  
Typical Data ◽  
Field Temperature

This paper reports and compares Raman and infrared thermometry measurements along the legs and on the shuttle of a SOI (silicon on insulator) bent-beam thermal microactuator. Raman thermometry offers micron spatial resolution and measurement uncertainties of ±10 K. Typical data collection times are a minute per location leading to measurement times on the order of hours for a complete temperature profile. Infrared thermometry obtains a full-field measurement so the data collection time is on the order of a minute. The spatial resolution is determined by the pixel size, 25 μm by 25 μm for the system used, and infrared thermometry also has uncertainties of ±10 K after calibration with a nonpackaged sample. The Raman and infrared measured temperatures agreed both qualitatively and quantitatively. For example, when the thermal microactuator was operated at 7 V, the peak temperature on an interior leg is 437 K ± 10 K and 433 K ± 10 K from Raman and infrared thermometry, respectively. The two techniques are complementary for microsystems characterization when infrared imaging obtains a full-field temperature measurement and Raman thermometry interrogates regions for which higher spatial resolution is required.

Benchmarking driving efficiency using data science techniques applied on large - scale smartphone data

2018 ◽  
Author(s):  
Δημήτριος Τσελέντης
Keyword(s):  
Data Envelopment Analysis ◽  
Convex Hull ◽  
Large Scale ◽  
Data Science ◽  
Data Envelopment ◽  
Using Data

Ο κύριος στόχος της παρούσας διδακτορικής διατριβής είναι η ανάπτυξη μιας ολοκληρωμένης μεθοδολογικής προσέγγισης για τη συγκριτική αξιολόγηση της οδηγικής επίδοσης, όσον αφορά την οδική ασφάλεια, τόσο σε επίπεδο διαδρομής, όσο και οδηγού, με τη χρήση τεχνικών της επιστήμης δεδομένων. Η μεθοδολογική προσέγγιση στηρίζεται στον καθορισμό ενός δείκτη επίδοσης που βασίζεται στη θεωρία της Περιβάλλουσας Ανάλυσης Δεδομένων (Data Envelopment Analysis - DEA) και σχετίζεται με μακροσκοπικά συμπεριφοριστικά χαρακτηριστικά οδήγησης, όπως ο αριθμός των απότομων επιταχύνσεων/ επιβραδύνσεων, ο χρόνος χρήσης του κινητού τηλεφώνου και ο χρόνος υπέρβασης του ορίου ταχύτητας. Ακόμα, αναπτύσσονται μοντέλα μηχανικής μάθησης για τον προσδιορισμό διακριτών προφίλ οδήγησης που βασίζονται στη χρονική εξέλιξη της οδηγικής επίδοσης. Η προτεινόμενη μεθοδολογική προσέγγιση εφαρμόζεται σε πραγματικά δεδομένα οδήγησης ευρείας κλίμακας που συλλέγονται από έξυπνες συσκευές κινητών τηλεφώνων (smartphones), τα οποία αναλύονται μέσω στατιστικών μεθόδων για τον προσδιορισμό της απαιτούμενης ποσότητας δεδομένων οδήγησης που θα χρησιμοποιηθούν στην ανάλυση. Τα αποτελέσματα δείχνουν ότι ο βελτιστοποιημένος αλγόριθμος convex hull – DEA δίνει εξίσου ακριβή και ταχύτερα αποτελέσματα σε σχέση με τις κλασικές προσεγγίσεις της DEA. Ακόμα, η μεθοδολογία επιτρέπει τον προσδιορισμό των λιγότερο αποδοτικών ταξιδιών σε μια βάση δεδομένων καθώς και το αποδοτικό επίπεδο οδηγικών στοιχείων ενός ταξιδιού για να καταστεί αποδοτικότερη από την άποψη της ασφάλειας. Η περαιτέρω ομάδοποίηση των οδηγών με βάση της απόδοσή τους σε βάθος χρόνου οδηγεί στον εντοπισμό τριών ομάδων οδηγών, αυτή του μέσου οδηγού, του ασταθή οδηγού και του λιγότερο επικίνδυνου οδηγού. Τα αποτελέσματα δείχνουν ότι η εκ των προτέρων γνώση σχετικά με το ιστορικό ατυχημάτων του χρήστη φαίνεται να επηρεάζουν μόνο τη σύσταση της δεύτερης συστάδας των πιο ασταθών οδηγών, η οποία ενσωματώνει τους οδηγούς που είναι λιγότερο αποδοτικοί και ασταθής ως προς την ασφάλεια. Φαίνεται επίσης ότι η χρήση κινητών τηλεφώνων δεν αποτελεί κρίσιμο παράγοντα για τον καθορισμό της επίδοσης της ασφάλειας ενός οδηγού, καθώς διαπιστώθηκαν μικρές διαφορές σε σχέση με αυτό το χαρακτηριστική οδήγησης μεταξύ οδηγών διαφορετικών κατηγοριών επίδοσης. Επιπλέον, δείχνεται ότι απαιτείται μια διαφορετική δειγματοληψίας δεδομένων οδήγησης για κάθε α) οδικό τύπο, β) χαρακτηριστικό οδήγησης και γ) οδηγική επιθετικότητα για να συγκεντρωθούν αρκετά δεδομένα και να αποκτηθεί μια σαφής εικόνα της οδηγικής συμπεριφοράς και να εκτελεστεί ανάλυση με χρήση DEA. Τα αποτελέσματα θα μπορούσαν να αξιοποιηθούν για την παροχή εξατομικευμένης ανατροφοδότησης στους οδηγούς σχετικά με τη συνολική τους οδηγική επίδοση και την εξέλιξή της, προκειμένου να βελτιωθεί και να μειωθεί ο κίνδυνος ατυχήματος.

Using of data science in healthcare

2021 ◽  
pp. 149-156
Author(s):  
Ihor Ponomarenko ◽  
Oleksandra Lubkovska
Keyword(s):  
Machine Learning ◽  
Health Care ◽  
Business Intelligence ◽  
Data Science ◽  
Large Data ◽  
Science Methods ◽  
Medical Field ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Using Data

The subject of the research is the approach to the possibility of using data science methods in the field of health care for integrated data processing and analysis in order to optimize economic and specialized processes The purpose of writing this article is to address issues related to the specifics of the use of Data Science methods in the field of health care on the basis of comprehensive information obtained from various sources. Methodology. The research methodology is system-structural and comparative analyzes (to study the application of BI-systems in the process of working with large data sets); monograph (the study of various software solutions in the market of business intelligence); economic analysis (when assessing the possibility of using business intelligence systems to strengthen the competitive position of companies). The scientific novelty the main sources of data on key processes in the medical field. Examples of innovative methods of collecting information in the field of health care, which are becoming widespread in the context of digitalization, are presented. The main sources of data in the field of health care used in Data Science are revealed. The specifics of the application of machine learning methods in the field of health care in the conditions of increasing competition between market participants and increasing demand for relevant products from the population are presented. Conclusions. The intensification of the integration of Data Science in the medical field is due to the increase of digitized data (statistics, textual informa- tion, visualizations, etc.). Through the use of machine learning methods, doctors and other health professionals have new opportunities to improve the efficiency of the health care system as a whole. Key words: Data science, efficiency, information, machine learning, medicine, Python, healthcare.

An Assessment on Classification in Python Using Data Science

2021 ◽  
Author(s):  
Margaret Mary T ◽  
Soumya K ◽  
Ramanathan G ◽  
Clinton G
Keyword(s):  
Data Science ◽  
Using Data
Sign in / Sign up

Export Citation Format

Share Document