首页|New Machine Learning Study Findings Have Been Reported by Researchers at Univers ity of Milan (On the Robustness of Random Forest Against Untargeted Data Poisoni ng: an Ensemble-based Approach)
New Machine Learning Study Findings Have Been Reported by Researchers at Univers ity of Milan (On the Robustness of Random Forest Against Untargeted Data Poisoni ng: an Ensemble-based Approach)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Research findings on Machine Learning are discussed in a new report. According to news reporting from Milan, Italy, by NewsRx journalists, research stated, "Machine learning is becoming ubiquitous. From finance to medicine, machine learning models are boosting decision-making p rocesses and even outperforming humans in some tasks." Financial support for this research came from Technology Innovation Institute. The news correspondents obtained a quote from the research from the University o f Milan, "This huge progress in terms of prediction quality does not however fin d a counterpart in the security of such models and corresponding predictions, wh ere perturbations of fractions of the training set (poisoning) can seriously und ermine the model accuracy. Research on poisoning attacks and defenses received i ncreasing attention in the last decade, leading to several promising solutions a iming to increase the robustness of machine learning. Among them, ensemble-based defenses, where different models are trained on portions of the training set an d their predictions are then aggregated, provide strong theoretical guarantees a t the price of a linear overhead. Surprisingly, ensemble-based defenses, which d o not pose any restrictions on the base model, have not been applied to increase the robustness of random forest. The work in this paper aims to fill in this ga p by designing and implementing a novel hash-based ensemble approach that protec ts random forest against untargeted, random poisoning attacks. An extensive expe rimental evaluation measures the performance of our approach against a variety o f attacks, as well as its sustainability in terms of resource consumption and pe rformance, and compares it with a traditional monolithic model based on random f orest."
MilanItalyEuropeCyborgsEmerging TechnologiesHealth and MedicineMachine LearningPoisoningUniversity of Mi lan
NETL
NSTL
