Privacy and security-related concerns are growing as machine learning reaches diverse application domains. The data holders want to train or infer with private data while exploiting accelerators, such as GPUs, that are hosted in the cloud. Cloud systems are vulnerable to attackers that compromise the privacy of data and integrity of computations. Tackling such a challenge requires unifying theoretical privacy algorithms with hardware security capabilities. This paper presents DarKnight, a framework for large DNN training while protecting input privacy and computation integrity. DarKnight relies on cooperative execution between trusted execution environments (TEE) and accelerators, where the TEE provides privacy and integrity verification, while accelerators perform the bulk of the linear algebraic computation to optimize the performance. In particular, DarKnight uses a customized data encoding strategy based on matrix masking to create input obfuscation within a TEE. The obfuscated data is then offloaded to GPUs for fast linear algebraic computation. DarKnight's data obfuscation strategy provides provable data privacy and computation integrity in the cloud servers. While prior works tackle inference privacy and cannot be utilized for training, DarKnight's encoding scheme is designed to support both training and inference.

Speech emotion recognition (SER) processes speech signals to detect and characterize expressed perceived emotions. Many SER application systems often acquire and transmit speech data collected at the client-side to remote cloud platforms for inference and decision making. However, speech data carry rich information not only about emotions conveyed in vocal expressions, but also other sensitive demographic traits such as gender, age and language background. Consequently, it is desirable for SER systems to have the ability to classify emotion constructs while preventing unintended/improper inferences of sensitive and demographic information. Federated learning (FL) is a distributed machine learning paradigm that coordinates clients to train a model collaboratively without sharing their local data. This training approach appears secure and can improve privacy for SER. However, recent works have demonstrated that FL approaches are still vulnerable to various privacy attacks like reconstruction attacks and membership inference attacks. Although most of these have focused on computer vision applications, such information leakages exist in the SER systems trained using the FL technique. To assess the information leakage of SER systems trained using FL, we propose an attribute inference attack framework that infers sensitive attribute information of the clients from shared gradients or model parameters, corresponding to the FedSGD and the FedAvg training algorithms, respectively. As a use case, we empirically evaluate our approach for predicting the client's gender information using three SER benchmark datasets: IEMOCAP, CREMA-D, and MSP-Improv. We show that the attribute inference attack is achievable for SER systems trained using FL. We further identify that most information leakage possibly comes from the first layer in the SER model.

No abstract

Stragglers, Byzantine workers, and data privacy are the main bottlenecks in distributed cloud computing. Several prior works proposed coded computing strategies to jointly address all three challenges. They require either a large number of workers, a significant communication cost or a significant computational complexity to tolerate malicious workers. Much of the overhead in prior schemes comes from the fact that they tightly couple coding for all three problems into a single framework. In this work, we propose Verifiable Coded Computing (VCC) framework that decouples Byzantine node detection challenge from the straggler tolerance. VCC leverages coded computing just for handling stragglers and privacy, and then uses an orthogonal approach of verifiable computing to tackle Byzantine nodes. Furthermore, VCC dynamically adapts its coding scheme to tradeoff straggler tolerance with Byzantine protection and vice-versa. We evaluate VCC on compute intensive distributed logistic regression application. Our experiments show that VCC speeds up the conventional uncoded implementation of distributed logistic regression by 3.2 × −6.9×, and also improves the test accuracy by up to 12.6%.

Context: Student assessment is an essential part of higher education. Many different technology-based assessment methods have been formed with the increasing development of IT and its introduction into the education system. Online take-home exams are computer-based exams in which the examinees can take at a place of their own choice and on their own computers. Despite its benefits, this method is faced with certain problems. The present study investigates the challenges in holding take-home computer-based exams in medical sciences and various solutions proposed to use this method more extensively in Iran in situations of crisis. Evidence Acquisition: The present review article was drafted upon a search conducted in Scopus, Google Scholar, and Google’s general search engine using the following keywords and search strategies: "Take-home exam", OR "Take-home assessment", OR, "Online exam", OR "Online assessment", AND "Higher education". The content of the related documents published from 2009 to 2020, including articles, books, and web pages, was selected and assessed, and 35 articles were finally used to accomplish the study objectives. Results: Online take-home exams have many advantages, including reduced human errors, rapid scoring, and reduced stress on the examinees. Nonetheless, one of the limitations of this examination method is that the examinees may not meet all the criteria required for taking exams at home. The obvious risk is students’ unethical conduct and cheating, which composes a major challenge of this examination modality. Conclusions: The reliability and correctness of exams can be improved using combination techniques, question banks, and giving random equivalent questions to each candidate that are not necessarily similar, and also mixing up the questions and their answers, which can provide a tool for preventing or limiting cheating. Online monitoring systems are also one of the strategies proposed for ongoing monitoring of online exams by an invigilator that are generally developed through artificial intelligence.

No abstract

No abstract

Predicting student academic performance in educational information systems becomes one of the major concerns in improving the quality of academic institutions. Educational data mining can identify the settings that characterize students’ behavior. This study develops prediction models for students’ performance using neural networks, deep learning, and random forest. Deep learning uses multiple hidden layers to represent the data at a higher level, using linear or non-linear transformations. Random Forest that is, a variety of Decision tree algorithms uses the randomization concept to produce many models based on a random selection of input space. We train the models on a national exam dataset. In the following, we compare results with the neural network as a well-known method. The experimental result shows that Deep learning finds a deep data structure by finding better features and achieves accurate results. Although, Random Forest gains a higher accuracy in predicting students’ performance.