Di Zhuang

The generalization capability of machine learning models, which refers to generalizing the knowledge for an ``unseen'' domain via learning from one or multiple seen domain(s), is of great importance to develop and deploy machine learning applications in the real-world conditions. Domain Generalization (DG) techniques aim to enhance such generalization capability of machine learning models, where the learnt feature representation and the classifier are two crucial factors to improve… Show more

The generalization capability of machine learning models, which refers to generalizing the knowledge for an ``unseen'' domain via learning from one or multiple seen domain(s), is of great importance to develop and deploy machine learning applications in the real-world conditions. Domain Generalization (DG) techniques aim to enhance such generalization capability of machine learning models, where the learnt feature representation and the classifier are two crucial factors to improve generalization and make decisions. In this paper, we propose Discriminative Adversarial Domain Generalization (DADG) with meta-learning-based cross-domain validation. Our proposed framework contains two main components that work synergistically to build a domain-generalized DNN model: (i) discriminative adversarial learning, which proactively learns a generalized feature representation on multiple ``seen'' domains, and (ii) meta-learning based cross domain validation, which simulates train/test domain shift via applying meta-learning techniques in the training process. In the experimental evaluation, a comprehensive comparison has been made among our proposed approach and other existing approaches on three benchmark datasets. The results shown that DADG consistently outperforms a strong baseline DeepAll, and outperforms the other existing DG algorithms in most of the evaluation cases. Show less

In the big data era, more and more cloud-based data-driven applications are developed that leverage individual data to provide certain valuable services (the utilities). On the other hand, since the same set of individual data could be utilized to infer the individual's certain sensitive information, it creates new channels to snoop the individual's privacy. Hence it is of great importance to develop techniques that enable the data owners to release privatized data, that can still be utilized… Show more

In the big data era, more and more cloud-based data-driven applications are developed that leverage individual data to provide certain valuable services (the utilities). On the other hand, since the same set of individual data could be utilized to infer the individual's certain sensitive information, it creates new channels to snoop the individual's privacy. Hence it is of great importance to develop techniques that enable the data owners to release privatized data, that can still be utilized for certain premised intended purpose. Existing data releasing approaches, however, are either privacy-emphasized (no consideration on utility) or utility-driven (no guarantees on privacy). In this work, we propose a two-step perturbation-based utility-aware privacy-preserving data releasing framework. First, certain predefined privacy and utility problems are learned from the public domain data (background knowledge). Later, our approach leverages the learned knowledge to precisely perturb the data owners' data into privatized data that can be successfully utilized for certain intended purpose (learning to succeed), without jeopardizing certain predefined privacy (training to fail). Extensive experiments have been conducted on Human Activity Recognition, Census Income and Bank Marketing datasets to demonstrate the effectiveness and practicality of our framework. Show less

Convolutional neural networks (CNNs) have achieved the state-of-the-art performance in skin lesion analysis. Compared with single CNN classifier, combining the results of multiple classifiers via fusion approaches shows to be more effective and robust. Since the skin lesion datasets are usually limited and statistically biased, while designing an effective fusion approach, it is important to consider not only the performance of each classifier on the training/validation dataset, but also the… Show more

Convolutional neural networks (CNNs) have achieved the state-of-the-art performance in skin lesion analysis. Compared with single CNN classifier, combining the results of multiple classifiers via fusion approaches shows to be more effective and robust. Since the skin lesion datasets are usually limited and statistically biased, while designing an effective fusion approach, it is important to consider not only the performance of each classifier on the training/validation dataset, but also the relative discriminative power (e.g., confidence) of each classifier regarding an individual sample in the testing phase, which calls for an active fusion approach. Furthermore, in skin lesion analysis, the data of certain classes is usually abundant making them an over-represented majority (e.g., benign lesions), while the data of some other classes is deficient, making them an underrepresented minority (e.g., cancerous lesions). It is more crucial to precisely identify the samples from an underrepresented (i.e., in terms of the amount of data) but more important (e.g., the cancerous lesions) minority class. In other words, misclassifying a more severe lesion to a benign or less severe lesion should have relative more cost (e.g., money, time and even lives). To address such challenges, we present CS-AF, a cost-sensitive multi-classifier active fusion framework for skin lesion classification. In the experimental evaluation, we prepared 60 base classifiers (of 10 CNN architectures) on the ISIC research datasets. Our experimental results show that our framework consistently outperforms the static fusion competitors. Show less

As the advancement of deep learning (DL), the Internet of Things and cloud computing techniques for biomedical and healthcare problems, mobile healthcare systems have received unprecedented attention. Since DL techniques usually require enormous amount of computation, most of them cannot be directly deployed on the resource-constrained mobile and IoT devices. Hence, most of the mobile healthcare systems leverage the cloud computing infrastructure, where the data collected by the mobile and IoT… Show more

As the advancement of deep learning (DL), the Internet of Things and cloud computing techniques for biomedical and healthcare problems, mobile healthcare systems have received unprecedented attention. Since DL techniques usually require enormous amount of computation, most of them cannot be directly deployed on the resource-constrained mobile and IoT devices. Hence, most of the mobile healthcare systems leverage the cloud computing infrastructure, where the data collected by the mobile and IoT devices would be transmitted to the cloud computing platforms for analysis. However, in the contested environments, relying on the cloud might not be practical at all times. For instance, the satellite communication might be denied or disrupted. We propose SAIA, a Split Artificial Intelligence Architecture for mobile healthcare systems. Unlike traditional approaches for artificial intelligence (AI) which solely exploits the computational power of the cloud server, SAIA could not only relies on the cloud computing infrastructure while the wireless communication is available, but also utilizes the lightweight AI solutions that work locally on the client side, hence, it can work even when the communication is impeded. In SAIA, we propose a meta-information based decision unit, that could tune whether a sample captured by the client should be operated by the embedded AI (i.e., keeping on the client) or the networked AI (i.e., sending to the server), under different conditions. In our experimental evaluation, extensive experiments have been conducted on two popular healthcare datasets. Our results show that SAIA consistently outperforms its baselines in terms of both effectiveness and efficiency. Show less

Protecting sensitive information against data exploiting attacks is an emerging research area in data mining. Over the past, several different methods have been introduced to protect individual privacy from such attacks while maximizing data-utility of the application. However, these existing techniques are not sufficient to effectively protect data owner privacy, especially in the scenarios that utilize visualizable data (e.g. images, videos) or the applications that require heavy computations… Show more

Protecting sensitive information against data exploiting attacks is an emerging research area in data mining. Over the past, several different methods have been introduced to protect individual privacy from such attacks while maximizing data-utility of the application. However, these existing techniques are not sufficient to effectively protect data owner privacy, especially in the scenarios that utilize visualizable data (e.g. images, videos) or the applications that require heavy computations for implementation. To address these problems, we propose a new dimension reduction-based method for privacy preservation. Our method generates dimension-reduced data for performing machine learning tasks and prevents a strong adversary from reconstructing the original data. We first introduce a theoretical approach to evaluate dimension reduction-based privacy preserving mechanisms, then propose a non-linear dimension reduction framework motivated by state-of-the-art neural network structures for privacy preservation. We conducted experiments over three different face image datasets (AT&T, YaleB, and CelebA), and the results show that when the number of dimensions is reduced to seven, we can achieve the accuracies of 79%, 80%, and 73% respectively and the reconstructed images are not recognizable to naked human eyes. Show less

Community detection is of great importance for online social network analysis. The volume, variety and velocity of data generated by today's online social networks are advancing the way researchers analyze those networks. For instance, real-world networks, such as Facebook, LinkedIn and Twitter, are inherently growing rapidly and expanding aggressively over time. However, most of the studies so far have been focusing on detecting communities on the static networks. It is computationally… Show more

Community detection is of great importance for online social network analysis. The volume, variety and velocity of data generated by today's online social networks are advancing the way researchers analyze those networks. For instance, real-world networks, such as Facebook, LinkedIn and Twitter, are inherently growing rapidly and expanding aggressively over time. However, most of the studies so far have been focusing on detecting communities on the static networks. It is computationally expensive to directly employ a well-studied static algorithm repeatedly on the network snapshots of the dynamic networks. We propose DynaMo, a novel modularity-based dynamic community detection algorithm, aiming to detect communities of dynamic networks as effective as repeatedly applying static algorithms but in a more efficient way. DynaMo is an adaptive and incremental algorithm, which is designed for incrementally maximizing the modularity gain while updating the community structure of dynamic networks. In the experimental evaluation, a comprehensive comparison has been made among DynaMo, Louvain (static) and 5 other dynamic algorithms. Extensive experiments have been conducted on 6 real-world networks and 10,000 synthetic networks. Our results show that DynaMo outperforms all the other 5 dynamic algorithms in terms of the effectiveness, and is 2 to 5 times (by average) faster than Louvain algorithm. Show less

Peer-to-peer (P2P) botnets have become one of the major threats in network security for serving as the fundamental infrastructure for various cyber-crimes. More challenges are involved in the problem of detecting P2P botnets, despite a few work claimed to detect centralized botnets effectively. We propose Enhanced PeerHunter, a network-flow level community behavior analysis based system, to detect P2P botnets. Our system starts from a P2P network flow detection component. Then, it uses “mutual… Show more

Peer-to-peer (P2P) botnets have become one of the major threats in network security for serving as the fundamental infrastructure for various cyber-crimes. More challenges are involved in the problem of detecting P2P botnets, despite a few work claimed to detect centralized botnets effectively. We propose Enhanced PeerHunter, a network-flow level community behavior analysis based system, to detect P2P botnets. Our system starts from a P2P network flow detection component. Then, it uses “mutual contacts” to cluster bots into communities. Finally, it uses network-flow level community behavior analysis to detect potential botnets. In the experimental evaluation, we propose two evasion attacks, where we assume the adversaries know our techniques in advance and attempt to evade our system by making the P2P bots mimic the behavior of legitimate P2P applications. Our results showed that Enhanced PeerHunter can obtain high detection rate with few false positives, and high robustness against the proposed attacks. Show less

Data-driven mobile applications are becoming increasingly popular in civilian and law enforcement. RapidGather, for instance, is an smartphone application that collects data from individual, and spreads rapid emergency responses. Image data is widely used in such applications, and machine learning methods could be utilized to analyze the image data. However, people would hesitate to share the data without protecting their privacy. In this paper, we propose to utilize dimensionality reduction… Show more

Data-driven mobile applications are becoming increasingly popular in civilian and law enforcement. RapidGather, for instance, is an smartphone application that collects data from individual, and spreads rapid emergency responses. Image data is widely used in such applications, and machine learning methods could be utilized to analyze the image data. However, people would hesitate to share the data without protecting their privacy. In this paper, we propose to utilize dimensionality reduction techniques for privacy-preserving machine learning in face recognition for the image data. To demonstrate the proposed approach, we implement a client server system, FRiPAL. With extensive experiments, we show that FRiPAL is efficient, and could preserve the privacy of data owners while maintaining the utility for data users. Show less

Peer-to-peer (P2P) botnets have become one of the major threats in network security for serving as the infrastructure that responsible for various of cyber-crimes. Though a few existing work claimed to detect traditional botnets effectively, the problem of detecting P2P botnets involves more challenges. In this paper, we present PeerHunter, a community behavior analysis based method, which is capable of detecting botnets that communicate via a P2P structure. PeerHunter starts from a P2P hosts… Show more

Peer-to-peer (P2P) botnets have become one of the major threats in network security for serving as the infrastructure that responsible for various of cyber-crimes. Though a few existing work claimed to detect traditional botnets effectively, the problem of detecting P2P botnets involves more challenges. In this paper, we present PeerHunter, a community behavior analysis based method, which is capable of detecting botnets that communicate via a P2P structure. PeerHunter starts from a P2P hosts detection component. Then, it uses mutual contacts as the main feature to cluster bots into communities. Finally, it uses community behavior analysis to detect potential botnet communities and further identify bot candidates. Through extensive experiments with real and simulated network traces, PeerHunter can achieve very high detection rate and low false positives. Show less