Ying Zhao


1 Learning User Keystroke Patterns for Authentication

Authors: Ying Zhao


Keystroke authentication is a new access control system to identify legitimate users via their typing behavior. In this paper, machine learning techniques are adapted for keystroke authentication. Seven learning methods are used to build models to differentiate user keystroke patterns. The selected classification methods are Decision Tree, Naive Bayesian, Instance Based Learning, Decision Table, One Rule, Random Tree and K-star. Among these methods, three of them are studied in more details. The results show that machine learning is a feasible alternative for keystroke authentication. Compared to the conventional Nearest Neighbour method in the recent research, learning methods especially Decision Tree can be more accurate. In addition, the experiment results reveal that 3-Grams is more accurate than 2-Grams and 4-Grams for feature extraction. Also, combination of attributes tend to result higher accuracy.

Keywords: Pattern Recognition, Bayesian, Decision Tree, machinelearning, Keystroke Authentication, Instance-based Learning

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2511


3 Biodistribution of Fluorescence-Labelled Epidermal Growth Factor Protein from Slow Release Nanozolid Depots in Mouse

Authors: Stefan Gruden, Charlott Brunmark, Bo Holmqvist, Erwin D. Brenndorfer, Martin Johansson, Jian Liu, Ying Zhao, Niklas Axen, Moustapha Hassan


Aim: The study was designed to evaluate the ability of the calcium sulfate-based NanoZolid® drug delivery technology to locally release the epidermal growth factor (EGF) protein while maintaining its biological activity. Methods: NanoZolid-formulated EGF protein labelled with a near-infrared dye (EGF-NIR) depots or EGF-NIR dissolved in PBS were injected subcutaneously into mice bearing EGF receptor (EGFR) positive human A549 lung cancer tumors inoculated subcutaneously. The release and biodistribution of the EGF-NIR were investigated in vivo longitudinally up to 96 hours post-administration, utilizing whole-body fluorescence imaging. In order to confirm the in vivo findings, histological analysis of tumor cryosections was performed to investigate EGF-NIR fluorescent signal and EGFR expression level by immunofluorescence labelling. Results: The in vivo fluorescence imaging showed a controlled release profile of the EGF-NIR loaded in the NanoZolid depots compared to free EGF-NIR. Histological analysis of the tumors further demonstrated a prevailing distribution of EGF-NIR in regions with high levels of EGFR expression. Conclusion: Calcium sulfate based depots can be used to formulate EGF while maintaining its biological activity, e.g., receptor binding capability. This may have good clinical potential for local delivery of biomolecules to enhance treatment efficacy and minimize systemic adverse effects.

Keywords: controlled release, bioresorbable, calcium sulfate, NanoZolid

Procedia PDF Downloads 1
2 Inactivation of Semicarbazide-Sensitive Amine Oxidase Induces the Phenotypic Switch of Smooth Muscle Cells and Aggravates the Development of Atherosclerotic Lesions

Authors: Miao Zhang, Limin Liu, Feng Zhi, Panpan Niu, Mengya Yang, Xuemei Zhu, Ying Diao, Jun Wang, Ying Zhao


Background and Aims: Clinical studies have demonstrated that serum semicarbazide-sensitive amine oxidase (SSAO) activities positively correlate with the progression of atherosclerosis. The aim of the present study is to investigate the effect of SSAO inactivation on the development of atherosclerosis. Methods: Female LDLr knockout (KO) mice were given the Western-type diet for 6 and 9 weeks to induce the formation of early and advanced lesions, and semicarbazide (SCZ, 0.125%) was added into the drinking water to inactivate SSAO in vivo. Results: Despite no impact on plasma total cholesterol levels, abrogation of SSAO by SCZ not only resulted in the enlargement of both early (1.5-fold, p=0.0043) and advanced (1.8-fold, p=0.0013) atherosclerotic lesions, but also led to reduced/increased lesion contents of macrophages/smooth muscle cells (SMCs) (macrophage: ~0.74-fold, p=0.0002(early)/0.0016(advanced); SMC: ~1.55-fold, p=0.0003(early) /0.0001(advanced)), respectively. Moreover, SSAO inactivation inhibited the migration of circulating monocytes into peripheral tissues and reduced the amount of circulating Ly6Chigh monocytes (0.7-fold, p=0.0001), which may account for the reduced macrophage content in lesions. In contrast, the increased number of SMCs in lesions of SCZ-treated mice is attributed to an augmented synthetic vascular SMC phenotype switch as evidenced by the increased proliferation of SMCs and accumulation of collagens in vivo. Conclusion: SSAO inactivation by SCZ promotes the phenotypic switch of SMCs and the development of atherosclerosis. The enzymatic activity of SSAO may thus represent a potential target in the prevention and/or treatment of atherosclerosis.

Keywords: Atherosclerosis, phenotype switch of smooth muscle cells, SSAO/VAP-1, semicarbazide

Procedia PDF Downloads 173
1 Semi-Supervised Hierarchical Clustering Given a Reference Tree of Labeled Documents

Authors: Ying Zhao, Xingyan Bin


Semi-supervised clustering algorithms have been shown effective to improve clustering process with even limited supervision. However, semi-supervised hierarchical clustering remains challenging due to the complexities of expressing constraints for agglomerative clustering algorithms. This paper proposes novel semi-supervised agglomerative clustering algorithms to build a hierarchy based on a known reference tree. We prove that by enforcing distance constraints defined by a reference tree during the process of hierarchical clustering, the resultant tree is guaranteed to be consistent with the reference tree. We also propose a framework that allows the hierarchical tree generation be aware of levels of levels of the agglomerative tree under creation, so that metric weights can be learned and adopted at each level in a recursive fashion. The experimental evaluation shows that the additional cost of our contraint-based semi-supervised hierarchical clustering algorithm (HAC) is negligible, and our combined semi-supervised HAC algorithm outperforms the state-of-the-art algorithms on real-world datasets. The experiments also show that our proposed methods can improve clustering performance even with a small number of unevenly distributed labeled data.

Keywords: semi-supervised clustering, hierarchical agglomerative clustering, reference trees, distance constraints

Procedia PDF Downloads 380