vishwasprasanna Goto Github PK

followers: 1.0 following: 2.0 repos: 7.0 gists: 0.0

Type: User

vishwasprasanna's Projects

3d_berzerk_arcade_game

Final Project from Computer Graphics I course by Dr. Ben Watson.

adbi

All the Projects from Algorithms and Data Guided Business Intelligence Course by Dr. Nagiza Samatova

Federated learning is inherently vulnerable to having the integrity of the global model compromised because the training data from which the model parameter updates have been derived (if they were not somehow artificially synthesized) is not available to evaluate the validity of the updates during the aggregation process. An adversary may attempt to poison the global model with updates that aim to weaken the ability of the model to classify accurately. In order to protect against such attacks, the various possible types of attacks must be enumerated, their most probable effects on the model updates identified, and appropriate countermeasures put in place to minimize the likelihood that such updates will be aggregated into the global model while maximizing the likelihood that at least a minimal proportion of legitimate updates will be accepted. In this work we explore these issues by simulating a visual federated learning environment that is being attacked by one or more malicious agents performing two types of targeted attacks, i.e. attacks whose goal is the misclassification of a subset of images while more or less preserving the overall performance of the global model. We implemented a mechanism to detect anomalous model updates and prevent their inclusion in the global model and compared the performance of the global model after training with and without this mechanism enabled.

alda_mobile_money_transaction_fraud_detection

Detecting a fraudulent mobile money transaction is the focus of our work. As mobile transactions continue to increase, online fraud detection continues to become a bigger issue. Although fraud via smartphones is increasing at a faster pace than general PC/laptop based fraud, smartphones have the potential to become as secure a channel as the web through the use of advanced encryption and authentication technologies. By paying close attention to red flags and suspicious activities, you can avoid merchant services fraud. According to the 2018 Global fraud report, it is evident that out of the digital market place consumers 91% of customers use smartphone out of which 88% use for personal banking and it has been noted that 72% cite fraud as a growing concern over the past twelve months and 63% report higher level of fraudulent losses over that same period. One such activity is cited in the rise of Synthetic identities. Synthetic identities come from accounts not held by actual individuals, but by fabricated identities created to perpetuate fraud. These identities are created by combining the credentials and information of a mixed set of individuals to create a completely new ID. Criminals use this kind of technique to commit frauds in the area of healthcare, utility services and taxes. The research in the area of combatting such kind of frauds, motivates us to find a robust system to detect fraudulent transactions. Smartphones have been an easy target for fraudsters as it lacks the security level that other mobile devices have. Fraudsters know that it is generally easier to take over an account by phishing, spear phishing (targeting an individual) or Smishing (phishing via a mobile device), than to open a new account using a real or ‘synthetic’ identity, which is why the risk of account takeover is one of the most alarming trends in fraud.

vishwasprasanna Goto Github PK

vishwasprasanna's Projects

3d_berzerk_arcade_game

adbi

adbi_capstone_project

alda_mobile_money_transaction_fraud_detection

ase

cg

graphlevel-anomalydetection

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent