摘要:
A system for performing an assessment of the robustness and resilience of an examined original ML model against model extraction attacks, comprising a computerized device having at least one processor, which is adapted to train multiple candidate models MC with the external dataset D for each of the specified candidate learning algorithms α in Alg, where each candidate substitute model is trained on a subset of D corresponding to the evaluated ith query limit of the query budget constraint Q; evaluate the performance of each substitute model MC according to different evaluation methods ∈ Evaluation; calculate the robustness of each substitute model, where smaller difference or high agreement/similarity rate between the performance of the original model and the substitute model indicates that the original and substitute models are similar to each other, and that the substitute model having the highest performance can mimic the behavior of the original model and can be used as a replica of the original model.
摘要:
A system for testing Machine Learning (ML) and deep learning models for robustness, and durability against adversarial bias and privacy attacks, comprising a Project Repository for storing metadata of ongoing projects each of which having a defined project policy, and created ML models and data sources being associated with the ongoing projects; a Secure Data Repository, for storing training and testing datasets and models used in each project for evaluating the robustness of the each project; a Data/Model Profiler for creating a profile, based on the settings and configurations of the datasets and the models; a Test Recommendation Engine for recommending the relevant and most indicative attacks/tests for each examined model and for creating indicative and effective test suites; a Test/Attack Ontology module for storing all attacks/tests with their metadata and mapping the attacks/tests to their corresponding settings and configurations; an Attack Repository for storing the implemented tests/attacks. An ML model is tested against each one of the robustness categories (privacy, bias and adversarial learning); a Test Execution Environment for Initializing a test suite, running multiple tests and prioritizing tests in the test suite; a Project/Test Analytics module for analyzing the test suite results and monitoring changes in performance over time; a Defenses Repository for storing implemented defense methods implemented for each robustness category.
摘要:
A system for computing an estimated energy consumption of a target program running on a specific computerized system, comprising a processor and associated memory, storing an application being run by the at least one processor, which are adapted to measure the energy consumption of every atomic action of the specific computerized system in different resolutions; create an energy cost for mapping between all atomic actions and their corresponding energy requirements for a specific computer system, on which the program will run; create a profile for the target program, in terms of which atomic actions of the specific computerized system are required for the execution of the target program on the specific computerized system; calculate the energy consumption of the target program, by calculating the amount of atomic actions related to the target program during its execution, multiplied by the energy consumption associated with every respective atomic action.
摘要:
A system for detecting suspicious websites in proxy's data streams, comprising a data collection and pre-processing module for receiving data from proxy logs and transforming the data into temporal website sequences of length n, where each sequence is from a specific user and removing rare websites that appear only once; a training module being a neural network for receiving each the sequence and performing a training phase, during which each sequence corresponds to a user ID and generating a language model for predicting the next token (website) in each sequence; an anomaly detection module for receiving all the sequences and feeding the sequences into the trained model; providing by the model, for every sequence, a probability score representing how probable the sequence is; classifying the sequence as suspicious if the score is above a specific threshold t; an alerting module having an alert-logic for outputting alerts based on the number of suspicious websites of a user and the number of websites that were blocked by the proxy.
摘要:
A system for automated neutralization of fileless malware on connected IoT devices, each having a memory, for storing the device's operation software, a processor for sending commands to the device's components and a network card, for connects the device's processor to a data network. The system comprises a Feature Extractor module for receiving, collecting and analyzing data from the device's memory, the processor and the network card, and for recording measurements from the device and extracting the device's behavioral pattern; an Intrusion Detector, for examining the behavioral pattern received from the Feature Extractor and deciding whether there was a malware break/attack or there is a malfunction of the device itself; a Remediation Selector, that learns in real-time how different pre-defined actions affect the monitored devices for receiving from the Intrusion Detector decision if the is device under a malware attack and its kind, and sending remediation, repairing and neutralization commands to an LSTM Neural Network.
摘要:
A method for automatically establishing multi-participant interactions and/or conference calls between users of mobile devices, according to which an interaction establishing application is installed on the mobile device of each participating user and receives for each user, phone numbers of other users defined as his friends and a list of predetermined triggers as conditions for initiating an interaction between users. The application constantly monitors the location of each user's mobile device and upon detecting that one or more of the triggers are met, the application automatically initiates one or more interactions between users that comply with the triggers that are met.
摘要:
A method for on demand matching between unused suppliers resources and consumers, comprising the steps of: implementing, on a computerized device (such as a server) with at least one processor, a resilient and trusted ad-hoc Peer-To-Peer (P2P) communication protocol across connected devices for sharing browsed content among an ad-hoc group of connected devices, using WiFi networking capabilities of the connected devices; optimizing, by the computerized device, internet consumption across browser based sessions, using standalone browsers, or browsers' modules utilized by native applications; replicating, by the computerized device, browser cache assets of participating devices, based on usage patterns and top internet destinations; refreshing, by the computerized device, shared cache assets in idle time for utilizing bandwidth across the ad hoc group; measuring, by the computerized device, internet experience optimization ratio over a predetermined period of time by monitoring the performance of devices participating in Home CDN vs. the devices that are not connected to a local network; measuring, by the computerized device, internet experience optimization ratio across multiple local networks setups.
摘要:
The invention is a method that enables a mobile communication service provider to estimate his market share on a street level in real time over a particular point of interest (POI). The POI can be any geographical region serviced by the provider and can range from whole countries down to groups of offices in a large office building. To deliver such output the method uses the network signaling data, which includes positioning data for active devices. Additionally, by using static data comprising the distribution of different models of devices employed by users in the network and the penetration rate of each model into the market, an accurate estimation of market share can be achieved.