Unorganized Malicious Attacks Detection
Authors: Ming Pang, Wei Gao, Min Tao, Zhi-Hua Zhou
NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we compare our proposed UMA with the state-of-the-art approaches for attack detection. We consider three common evaluating metrics for attack detection as in [13]: Precision = TP TP + FP, Recall = TP TP + FN, F1 = 2 Precision Recall Precision + Recall |
| Researcher Affiliation | Academia | National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, 210023, China |
| Pseudocode | Yes | Algorithm 1 The UMA Algorithm |
| Open Source Code | No | The paper does not provide any explicit statement about open-source code availability for the described methodology, nor does it include a link to a code repository. |
| Open Datasets | Yes | We first conduct our experiments on the common-used datasets Movie Lens100K and Movie Lens1M, released by Group Lens [25]. ... We also collect a real dataset Douban10K1 with attack profiles from Douban website, where registered users record rating information over various films, books, clothes, etc. We gather 12095 ratings of 213 users over 155 items. ... 1http://www.douban.com/. |
| Dataset Splits | No | The paper describes how attack profiles are generated and added to the datasets for evaluation (e.g., spam ratio, filler ratio), but does not specify explicit train/validation/test splits (e.g., percentages, sample counts, or cross-validation) for the rating matrix used by the UMA algorithm. |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments. |
| Software Dependencies | No | The paper does not list any specific software dependencies with version numbers (e.g., programming languages, libraries, or frameworks with their versions). |
| Experiment Setup | Yes | In the experiments, we set τ = 10/ m, α = 10/m and δ = p mn/200. A rating can be viewed as a malicious rating if it deviates from the ground-truth rating by more than 3, since the scale of ratings is from -2 to 2. We set parameter β = τ/3 according to Eq. (6) where the entries of Y will be nullified if they are smaller than the threshold. We set κ = τ under the convergence condition β (0, ( 33 5)κ/2) as in Theorem 2. |