scholarly journals Mitigating Location Privacy Attacks on Mobile Devices using Dynamic App Sandboxing

2019 ◽  
Vol 2019 (2) ◽  
pp. 66-87 ◽  
Author(s):  
Sashank Narain ◽  
Guevara Noubir
Keyword(s):  
Private Information ◽  
Location Privacy ◽  
Historical Data ◽  
Mobile Apps ◽  
Machine Learning Techniques ◽  
Traffic Information ◽  
Quadratic Program ◽  
Mobility Patterns ◽  
Learning Techniques ◽  
Location Inference

Abstract We present the design, implementation and evaluation of a system, called MATRIX, developed to protect the privacy of mobile device users from location inference and sensor side-channel attacks. MATRIX gives users control and visibility over location and sensor (e.g., Accelerometers and Gyroscopes) accesses by mobile apps. It implements a PrivoScope service that audits all location and sensor accesses by apps on the device and generates real-time notifications and graphs for visualizing these accesses; and a Synthetic Location service to enable users to provide obfuscated or synthetic location trajectories or sensor traces to apps they find useful, but do not trust with their private information. The services are designed to be extensible and easy for users, hiding all of the underlying complexity from them. MATRIX also implements a Location Provider component that generates realistic privacy-preserving synthetic identities and trajectories for users by incorporating traffic information using historical data from Google Maps Directions API, and accelerations using statistical information from user driving experiments. These mobility patterns are generated by modeling/solving user schedule using a randomized linear program and modeling/solving for user driving behavior using a quadratic program. We extensively evaluated MATRIX using user studies, popular location-driven apps and machine learning techniques, and demonstrate that it is portable to most Android devices globally, is reliable, has low-overhead, and generates synthetic trajectories that are difficult to differentiate from real mobility trajectories by an adversary.

scholarly journals Predicting COVID-19 Spread Level using Socio-Economic Indicators and Machine Learning Techniques

2020 ◽  
Author(s):  
Alaeddine Mihoub ◽  
Hosni Snoun ◽  
Moez Krichen ◽  
Riadh Bel Hadj Salah ◽  
Montassar Kahia
Keyword(s):  
Historical Data ◽  
Economic Indicators ◽  
Machine Learning Techniques ◽  
Factors Affecting ◽  
Novel Approach ◽  
Learning Techniques ◽  
Class Level ◽  
Feature Importance ◽  
F Measure ◽  
Relevant Factors

<p>The new so-called COVID-19 virus is unfortunately founded to be highly transmissible across the globe. In this study, we propose a novel approach for estimating the spread level of the virus for each country for three different dates between April and May 2020. Unlike previous studies, this investigation does not process any historical data of spread but rather relies on the socio-economic indicators of each country. Actually, more than 1000 socio-economic indicators and more than 190 countries were processed in this study. Concretely, data preprocessing techniques and feature selection approaches were applied to extract relevant indicators for the classification process. Countries around the globe were assigned to 4 classes of spread. To find the class level of each country, many classifiers were proposed based especially on Support Vectors Machines (SVM), Multi-Layer Perceptrons (MLP) and Random Forests (RF). Obtained results show the relevance of our approach since many classifiers succeeded in capturing the spread level, especially the RF classifier, with an F-measure equal to 93.85% for April 15th, 2020. Moreover, a feature importance study is conducted to deduce the best indicators to build robust spread level classifiers. However, as pointed out in the discussion, classifiers may face some difficulties for future dates since the huge increase of cases and the lack of other relevant factors affecting this widespread.<i></i></p>

scholarly journals Predicting COVID-19 Spread Level using Socio-Economic Indicators and Machine Learning Techniques

2020 ◽  
Author(s):  
Alaeddine Mihoub ◽  
Hosni Snoun ◽  
Moez Krichen ◽  
Riadh Bel Hadj Salah ◽  
Montassar Kahia
Keyword(s):  
Historical Data ◽  
Economic Indicators ◽  
Machine Learning Techniques ◽  
Factors Affecting ◽  
Novel Approach ◽  
Learning Techniques ◽  
Class Level ◽  
Feature Importance ◽  
F Measure ◽  
Relevant Factors

<p>The new so-called COVID-19 virus is unfortunately founded to be highly transmissible across the globe. In this study, we propose a novel approach for estimating the spread level of the virus for each country for three different dates between April and May 2020. Unlike previous studies, this investigation does not process any historical data of spread but rather relies on the socio-economic indicators of each country. Actually, more than 1000 socio-economic indicators and more than 190 countries were processed in this study. Concretely, data preprocessing techniques and feature selection approaches were applied to extract relevant indicators for the classification process. Countries around the globe were assigned to 4 classes of spread. To find the class level of each country, many classifiers were proposed based especially on Support Vectors Machines (SVM), Multi-Layer Perceptrons (MLP) and Random Forests (RF). Obtained results show the relevance of our approach since many classifiers succeeded in capturing the spread level, especially the RF classifier, with an F-measure equal to 93.85% for April 15th, 2020. Moreover, a feature importance study is conducted to deduce the best indicators to build robust spread level classifiers. However, as pointed out in the discussion, classifiers may face some difficulties for future dates since the huge increase of cases and the lack of other relevant factors affecting this widespread.<i></i></p>

Predicting Medical Resources Required to be Dispatched After Earthquake and Flood, Using Historical Data and Machine Learning Techniques

2018 ◽  
Vol 8 (2) ◽  
pp. 13-35
Author(s):  
Homer Papadopoulos ◽  
Antonis Korakis
Keyword(s):  
Machine Learning ◽  
Web Services ◽  
Large Scale ◽  
Historical Data ◽  
Machine Learning Techniques ◽  
Medical Resources ◽  
Simulation Tools ◽  
Medical Response ◽  
Emergency Medical ◽  
Learning Techniques

This article presents a method to predict the medical resources required to be dispatched after large-scale disasters to satisfy the demand. The historical data of past incidents (earthquakes, floods) regarding the number of victims requested emergency medical services and hospitalisation, simulation tools, web services and machine learning techniques have been combined. The authors adopted a twofold approach: a) use of web services and simulation tools to predict the potential number of victims and b) use of historical data and self-trained algorithms to “learn” from these data and provide relative predictions. Comparing actual and predicted victims needed hospitalisation showed that the proposed models can predict the medical resources required to be dispatched with acceptable errors. The results are promoting the use of electronic platforms able to coordinate an emergency medical response since these platforms can collect big heterogeneous datasets necessary to optimise the performance of the suggested algorithms.

scholarly journals Detecting Private Information in Large Social Network using mixed Machine Learning Techniques

2019 ◽  
Author(s):  
Pasquale De Luca
Keyword(s):  
Machine Learning ◽  
Social Network ◽  
Mobile Phones ◽  
Private Information ◽  
Current Problem ◽  
Personal Information ◽  
Private Life ◽  
Machine Learning Techniques ◽  
Learning Techniques ◽  
The Web

The violation of privacy, others people or personal, is a very current problem, which concerns not only on the web but also in private life. In the years 1990 it was expected that nowadays, that any routine operation was carried out "manually", and it would be performed through mobile phones or personal computers. The problem pertains the distribution network that allows to share and bring together information and as result the network becomes unsafe, if subjected to attacks. Nowaday we put personal information on web because otherwise we are seen as &ldquo;weak&rdquo;. This work aims to measure and analyze how much information are shared by users of a pre-established social network and it is carried out through a set of algorithms techniques of machine learning.

scholarly journals Gamified Mobile Applications for Improving Driving Behavior: A Systematic Mapping Study

2021 ◽  
Vol 2021 ◽  
pp. 1-24
Author(s):  
Abderrahim El hafidy ◽  
Taoufik Rachad ◽  
Ali Idri ◽  
Ahmed Zellou
Keyword(s):  
Machine Learning ◽  
Mobile Applications ◽  
Mobile Apps ◽  
Driving Behavior ◽  
Machine Learning Algorithms ◽  
Machine Learning Techniques ◽  
Systematic Mapping Study ◽  
Mapping Study ◽  
Systematic Mapping ◽  
Learning Techniques

Many research works and official reports approve that irresponsible driving behavior on the road is the main cause of accidents. Consequently, responsible driving behavior can significantly reduce accidents’ number and severity. Therefore, in the research area as well as in the industrial area, mobile technologies are widely exploited in assisting drivers in reducing accident rates and preventing accidents. For instance, several mobile apps are provided to assist drivers in improving their driving behavior. Recently and thanks to mobile cloud computing, smartphones can benefit from the computing power of servers in the cloud for executing machine learning algorithms. Therefore, many mobile applications of driving assistance and control are based on machine learning techniques to adjust their functioning automatically to driver history, context, and profile. Additionally, gamification is a key element in the design of these mobile applications that allow drivers to develop their engagement and motivation to improve their driving behavior. To have an overview concerning existing mobile apps that improve driving behavior, we have chosen to conduct a systematic mapping study about driving behavior mobile apps that exist in the most common mobile apps repositories or that were published as research works in digital libraries. In particular, we should explore their functionalities, the kinds of collected data, the used gamification elements, and the used machine learning techniques and algorithms. We have successfully identified 220 mobile apps that help to improve driving behavior. In this work, we will extract all the data that seem to be useful for the classification and analysis of the functionalities offered by these applications.

scholarly journals Network Intrusion Detection and Measuring the Data Set Performance by Machine Learning Technique (MLT)

2019 ◽  
Vol 8 (4) ◽  
pp. 11806-11809
Keyword(s):  
Machine Learning ◽  
Intrusion Detection ◽  
Computer Network ◽  
Detection System ◽  
Machine Learning Techniques ◽  
Traffic Information ◽  
Data Set ◽  
Identification Rate ◽  
Network Intrusion ◽  
Learning Techniques

Intrusion Detection System (IDS) is the most mainstream approach to protect a computer network from different malicious activities to identify an intrusion. There have been a lot of attempts towards more exceptional performance specifically in IDSs which depends on Data Mining (DM) and Machine Learning Techniques (MLT). Though there is a destructive issue in that available assessment, DataSet (DS), called KDD DS, can't reflect current network circumstances and the most recent attack situations. As far as we could know, there is no possible assessment DS. We present a novel evaluation DS in this paper, called Kyoto, based on the 5 years of actual traffic information, which derived from different sorts of honey pots. This Kyoto DS is utilized for testing and assessing distinctive MLT has examined in this work. The attention was on unprocessed measurements True +ve (TrPo), False +ve (FaPo), True – ve (TrNa), and False – ve (FaNa) to assess execution and to improve the identification rate of IDS.

Predicting Medical Resources Required to be Dispatched After Earthquake and Flood, Using Historical Data and Machine Learning Techniques

2020 ◽  
pp. 38-66
Author(s):  
Homer Papadopoulos ◽  
Antonis Korakis
Keyword(s):  
Machine Learning ◽  
Web Services ◽  
Large Scale ◽  
Historical Data ◽  
Machine Learning Techniques ◽  
Medical Resources ◽  
Simulation Tools ◽  
Medical Response ◽  
Emergency Medical ◽  
Learning Techniques

This article presents a method to predict the medical resources required to be dispatched after large-scale disasters to satisfy the demand. The historical data of past incidents (earthquakes, floods) regarding the number of victims requested emergency medical services and hospitalisation, simulation tools, web services and machine learning techniques have been combined. The authors adopted a twofold approach: a) use of web services and simulation tools to predict the potential number of victims and b) use of historical data and self-trained algorithms to “learn” from these data and provide relative predictions. Comparing actual and predicted victims needed hospitalisation showed that the proposed models can predict the medical resources required to be dispatched with acceptable errors. The results are promoting the use of electronic platforms able to coordinate an emergency medical response since these platforms can collect big heterogeneous datasets necessary to optimise the performance of the suggested algorithms.

Application of Machine Learning Techniques for the Analysis of National Bridge Inventory and Bridge Element Data

2019 ◽  
Vol 2673 (7) ◽  
pp. 99-110 ◽  
Author(s):  
Graziano Fiorillo ◽  
Hani Nassif
Keyword(s):  
Machine Learning ◽  
Asset Management ◽  
Historical Data ◽  
Machine Learning Techniques ◽  
The Past ◽  
Precise Evaluation ◽  
Learning Techniques ◽  
Good State ◽  
Wide Range ◽  
National Bridge Inventory

The MAP-21 Act requires information on bridge assets to be at the element level for management operations in the U.S.A. This approach has the objective of improving future predictions of the performance of bridge assets for a more precise evaluation of condition and correct allocation of management funds to keep bridges in a good state of repair. Although bridge element conditions were introduced in the 1990s, the application of such data had never been mandatory for bridge asset management until 2014, therefore, the amount of historical data on bridge element (BE) condition is still limited. On the other hand, National Bridge Inventory (NBI) ratings have been collected since the 1970s and a wide range of data are available. Therefore, it is natural to ask whether BE condition can be predicted using NBI data. In the past, researchers statistically related BE and NBI data, but little has been done to revert NBI to BE. This paper addresses both challenges of mapping BE–NBI condition data using several machine learning techniques. The results of the analysis of these techniques applied to a sample of about 9,000 bridges from northeastern states of the U.S.A. shows that between 79.8% and 100% of the NBI ratings for deck, superstructure, and substructure can be predicted within a rating error of ± 1. The back-mapping operation of NBI time-dependent ratings to BE deterioration profiles for deck, superstructure, and substructure can also be predicted accurately with a probability greater than 50% at the 95% confidence level.

scholarly journals Machine Learning Based Method for Prediction of Heart Disease in Big Data Environment

2020 ◽  
Vol 9 (4) ◽  
pp. 1917-1921
Keyword(s):  
Machine Learning ◽  
Big Data ◽  
Historical Data ◽  
Heart Diseases ◽  
Medical Professional ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Huge Data ◽  
Learning Techniques ◽  
Non Linear

Prediction of diseases is one of the challenging tasks in healthcare domain. Conventionally the heart diseases were diagnosed by experienced medical professional and cardiologist with the help of medical and clinical tests. With conventional method even experienced medical professional struggled to predict the disease with sufficient accuracy. In addition, manually analysing and extracting useful knowledge from the archived disease data becomes time consuming as well as infeasible. The advent of machine learning techniques enables the prediction of various diseases in healthcare domain. Machine learning algorithms are trained to learn from the existing historical data and prediction models are being created to predict the unknown raw data. For the past two decades, machine learning techniques are extensively employed for disease prediction. Despite the capability of machine algorithm on learning from huge historical data which is stored in data mart and data warehouses using traditional database technologies such as Oracle OnLine Analytical Processing (OLAP). The conventional database technologies suffer from the limitation that they cannot handle huge data or unstructured data or data that comes with speed. In this context, big data tools and technologies plays a major role in storing and facilitating the processing of huge data. In this paper, an approach is proposed for prediction of heart diseases using Support Vector Algorithm in Spark environment. Support Vector Machine algorithm is basically a binary classifier which classifies both linear and non-linear input data. It transforms the non-linear data into hyper plan with the help of different kernel functions. Spark is a distributed big data processing platform which has a unique feature of keeping and processing a huge data in memory. The proposed approach is tested with a benchmark dataset from UCI repository and results are discussed.

Sign in / Sign up

Export Citation Format

Share Document