PRID: Model Inversion Privacy Attacks in Hyperdimensional Learning Systems

2021 ◽  
Author(s):  
Alejandro Hernandez-Cano ◽  
Rosario Cammarota ◽  
Mohsen Imani
Keyword(s):  
Learning Systems ◽  
Model Inversion

scholarly journals Adversarial Learning of Privacy-Preserving and Task-Oriented Representations

2020 ◽  
Vol 34 (07) ◽  
pp. 12434-12441
Author(s):  
Taihong Xiao ◽  
Yi-Hsuan Tsai ◽  
Kihyuk Sohn ◽  
Manmohan Chandraker ◽  
Ming-Hsuan Yang
Keyword(s):  
Machine Learning ◽  
Privacy Protection ◽  
Input Data ◽  
Service Providers ◽  
Privacy Preserving ◽  
Learning Systems ◽  
Model Inversion ◽  
Perceptual Distance ◽  
Task Oriented ◽  
And Task

Data privacy has emerged as an important issue as data-driven deep learning has been an essential component of modern machine learning systems. For instance, there could be a potential privacy risk of machine learning systems via the model inversion attack, whose goal is to reconstruct the input data from the latent representation of deep networks. Our work aims at learning a privacy-preserving and task-oriented representation to defend against such model inversion attacks. Specifically, we propose an adversarial reconstruction learning framework that prevents the latent representations decoded into original input data. By simulating the expected behavior of adversary, our framework is realized by minimizing the negative pixel reconstruction loss or the negative feature reconstruction (i.e., perceptual distance) loss. We validate the proposed method on face attribute prediction, showing that our method allows protecting visual privacy with a small decrease in utility performance. In addition, we show the utility-privacy trade-off with different choices of hyperparameter for negative perceptual distance loss at training, allowing service providers to determine the right level of privacy-protection with a certain utility performance. Moreover, we provide an extensive study with different selections of features, tasks, and the data to further analyze their influence on privacy protection.

Assessment of Cognitive Load in Multimedia Learning Using Dual-Task Methodology

2002 ◽  
Vol 49 (2) ◽  
pp. 109-119 ◽  
Author(s):  
Roland Brünken ◽  
Susan Steinbacher ◽  
Jan L. Plass ◽  
Detlev Leutner
Keyword(s):  
Cognitive Load ◽  
Dual Task ◽  
Cognitive Load Theory ◽  
Multimedia Learning ◽  
Learning Systems ◽  
Learning Materials ◽  
New Approach ◽  
Direct Assessment ◽  
Load Theory ◽  
Audiovisual Presentation

Abstract. In two pilot experiments, a new approach for the direct assessment of cognitive load during multimedia learning was tested that uses dual-task methodology. Using this approach, we obtained the same pattern of cognitive load as predicted by cognitive load theory when applied to multimedia learning: The audiovisual presentation of text-based and picture-based learning materials induced less cognitive load than the visual-only presentation of the same material. The findings confirm the utility of dual-task methodology as a promising approach for the assessment of cognitive load induced by complex multimedia learning systems.

Platform for Analysing and Encouraging Student Activity on Contest and E-learning Systems

2018 ◽  
Vol 12 ◽  
pp. 85-98
Author(s):  
Bojan Kostadinov ◽  
Mile Jovanov ◽  
Emil STANKOV
Keyword(s):  
Machine Learning ◽  
Data Collection ◽  
Educational Policy ◽  
Learning Systems ◽  
Data Sources ◽  
Or Education ◽  
Student Activity ◽  
The World ◽  
E Learning ◽  
Analyse Data

Data collection and machine learning are changing the world. Whether it is medicine, sports or education, companies and institutions are investing a lot of time and money in systems that gather, process and analyse data. Likewise, to improve competitiveness, a lot of countries are making changes to their educational policy by supporting STEM disciplines. Therefore, it’s important to put effort into using various data sources to help students succeed in STEM. In this paper, we present a platform that can analyse student’s activity on various contest and e-learning systems, combine and process the data, and then present it in various ways that are easy to understand. This in turn enables teachers and organizers to recognize talented and hardworking students, identify issues, and/or motivate students to practice and work on areas where they’re weaker.

THE USE OF ICT IN ISLAMIC EDUCATION (PAI) LEARNING AT SCHOOL: A Case Study on Integrated Islamic High School (SMA-IT) Ummul Quro Bogor

2020 ◽  
Author(s):  
Farida Hanun
Keyword(s):  
High School ◽  
Learning Process ◽  
Research Method ◽  
Qualitative Approach ◽  
Learning Systems ◽  
Government Support ◽  
Islamic Education ◽  
Inhibiting Factors ◽  
The Impact

This study aims to obtain a description related to the learning of PAI by using ICT and how the impact of the use of ICT on PAI learning systems in the classroom. The research method uses a qualitative approach in the integrated Islamic high school Ummul Quro Bogor, West Java. The results showed that a) there were four stages of using ICT in the learning process, namely; emerging, applying, integrating dan transforming. PAI teachers are already at the integrating stage. In other words, ICT has been integrated into the PAI learning curriculum. b) supporting factors for the use of ICT are the existence of ICT support facilities, the availability of qualified educators, the commitment of the school to implement ICT in every PAI learning. c) Inhibiting factors in the use of ICT are aspects of financing ICT facilities require a large budget. Some elderly teachers have difficulty using ICT in the learning process. Besides, the internet network is unstable. d) The impact of the use of ICT is very significant on PAI learning process. e) the existence of ICT devices not only as a support but already as an important component in the education system. The research led to the recommendation of the need for government support in the form of concern for ICT in terms of policies, facilities, workforce, budget, and organizing training in the use of ICT for PAI teachers to improve their professionalism. Therefore, further research is suggested regarding the effectiveness of the use of ICT in the learning process of PAI.

Robot Teaching, Pedagogy, and Policy

2020 ◽  
pp. 787-803
Author(s):  
Elana Zeide
Keyword(s):  
Artificial Intelligence ◽  
Analytical Methods ◽  
Personalized Learning ◽  
Educational Process ◽  
Learning Systems ◽  
Policy Decisions ◽  
Public Discussion ◽  
Teaching Pedagogy ◽  
Do So

This chapter looks at the use of artificial intelligence (AI) in education, which immediately conjures the fantasy of robot teachers, as well as fears that robot teachers will replace their human counterparts. However, AI tools impact much more than instructional choices. Personalized learning systems take on a whole host of other educational roles as well, fundamentally reconfiguring education in the process. They not only perform the functions of robot teachers but also make pedagogical and policy decisions typically left to teachers and policymakers. Their design, affordances, analytical methods, and visualization dashboards construct a technological, computational, and statistical infrastructure that literally codifies what students learn, how they are assessed, and what standards they must meet. However, school procurement and implementation of these systems are rarely part of public discussion. If they are to remain relevant to the educational process itself, as opposed to just its packaging and context, schools and their stakeholders must be more proactive in demanding information from technology providers and setting internal protocols to ensure effective and consistent implementation. Those who choose to outsource instructional functions should do so with sufficient transparency mechanisms in place to ensure professional oversight guided by well-informed debate.

Sign in / Sign up

Export Citation Format

Share Document