Bo Yuan

Rochester Institute of Technology | RIT

As machine learning is deployed in more settings, including in security-sensitive applications such as malware detection, the risks posed by adversarial examples that fool machine-learning classifiers have become magnified. Black-box attacks are especially dangerous, as they only require the attacker to have the ability to query the target model an...

Machine learning models have been found to be susceptible to adversarial examples that are often indistinguishable from the original inputs. These adversarial examples are created by applying adversarial perturbations to input samples, which would cause them to be misclassified by the target models. Attacks that search and apply the perturbations t...

With the increasing number of attacks on enterprises, which often involves the deployment of some form of malware, an automated method for performing large-scale triage has become essential to the timely resolution of an incident. The purpose of this project is to combine the data collection capabilities of Google Rapid Response (GRR) with the flex...

Malware is the fastest growing threat to information technology systems. Although a single absolute solution for defeating malware is improbable, a stacked arsenal against malicious software enhances the ability to maintain security and privacy. This research attempts to reinforce the anti-malware arsenal by studying a behavioral activity common to...

Malware is one of the greatest and most rapidly growing threats to the digital world. Traditional signature-based detection is no longer adequate to detect new variants and highly targeted malware. Furthermore, dynamic detection is often circumvented with anti-VM and/or anti-debugger techniques. Recently heuristic approaches have been explored to e...

Accurate tracking of web attackers can be used to identify and monitor repeat offenders, and to allow defenders to deter and/or retaliate against the attackers. Tracking of web clients has been difficult since HTTP is a stateless protocol. Existing methods are limited to the current session and are not designed to track a client long-term or throug...

Identifying and attracting talented students to digital forensics programs is a crucial first step to developing professionals in this relatively young field. To respond to these challenges, we propose to develop a fun, entertaining, and yet educational forensicscourse suitable for first year students in college, in an effort to identify and attrac...

Research in quantitative Information Technology (IT) risk analysis has increased in the past decade, but much of that research has focused on creating new approaches that replace existing ones. Since organizations have extensive sunk costs invested in their risk management programs, there exists a need to extend and improve existing approaches. Add...

Covert communication is a rapidly expanding field of research with significant impact on the security theater. These communication methods, or “covert channels”, can be applied in a number of ways, including as a mechanism for an attacker to leak data from a monitored system or network. This paper sets out to contribute to this field by introducing...

We introduce a new system for layout-based (LATEX) indexing and retrieval of mathematical expressions using substitution trees. Substitution trees can efficiently store and find expressions based on the similarity of their symbols, symbol layout, sub-expressions and size. We describe our novel implementation and some of our modifications to the sub...

This paper presents a new covert channel based on Google Analytic web cookies in HTTP protocol. The new covert channel is difficult to disrupt and is capable of reasonably high bandwidths. The Google Analytic framework is used by over half of the most popular web sites currently on the Internet; its ubiquitousness across the web implies a great imp...

Two new methods for retrieving mathematical expressions using conventional keyword search and expression images are presented. An expression-level TF-IDF (term frequency-inverse document frequency) approach is used for keyword search, where queries and indexed expressions are represented by keywords taken from LATEX strings. TF-IDF is computed at t...

In the midst of a social networking revolution, social media has become the new vehicle for effective business marketing and transactions. As social aspects to the Internet continue to expand in both quantity and scope, so has the security threat towards enterprise networks and systems. Many social networking users also become main targets of spams...

A new covert channel over the RTP protocol is designed and implemented by modifying the timestamp value in the RTP header. Due to the high frequency of RTP packets, the covert channel has a high bit-rate, theoretically up to 350 bps. The broad use of RTP for multimedia applications such as VoIP, provides abundant opportunities to such a covert chan...

The ICMP protocol has been widely used and accepted as a covert channel. While the ICMP protocol is very simple to use, modern security approaches such as Firewalls, deep-packet inspection and intrusion detection systems threaten the use of ICMP for a reliable means for a covert channel. This study explores the modern usefulness of ICMP with typica...

Network covert channels allow two entities to communicate stealthily. Hypertext Transfer Protocol (HTTP), accounting for approximately half of all traffic on the Internet (Burke, 2007), has become the de facto standard for hiding network covert channels. Proliferation of covert channels throughout the World Wide Web has brought both challenges and...

Many covert channels take advantages of weaknesses, flaws, or unused data fields in network protocols In this paper, a behavior-based covert channel, that takes advantages of behavior of an application, is presented along with a formal definition in the framework of finite state machines The behavior-based covert channel is application specific and...

Underwater sensor deployment for military surveillance is a significant challenge due to the inherent difficulties posed by the underwater channel in terms of sensing and communications between sensors, as well as the exorbitant cost of the sensors. As a result, these sensors must be deployed as efficiently as possible. The proposed Underwater S...

Underwater sensor deployment for military surveillance is a significant challenge due to the inherent difficulties posed by the underwater channel in terms of sensing and communications between sensors, as well as the exorbitant cost of the sensors. As a result, these sensors must be deployed as efficiently as possible. The proposed Underwater S...

Underwater sensor deployment is a significant challenge due to the inherent difficulties posed by the underwater channel in terms of sensing and communications between sensors. In addition, monetary constraints arising from the cost of sensors and deploying them, limit the number of available sensors. As a result, these sensors must be deployed as...

As knowledge based systems become more sophisticated, communications between systems or among their subsystems often conducted over public channels such as the Internet, wireless medium, etc. To secure communications over public channels, the most often used method is Diffie and Hellman’s public key infrastructure approach. This method requires a t...

This paper presents a covert communication chan-nel that exists in virtually all forms of packet switching data net-works. On the one hand, this covert channel, if used properly, can potentially enhance the overall security of data communications over networks. On the other hand, the covert channel can also potentially become a back door to access...

In this paper, we propose several novel hardware circuits that can be used to design fuzzy logic systems. First, five different fuzzifying circuits, which can be easily embedded into other systems, are presented. Then, two novel defuzzifying schemes based on voltage and charge modes are introduced. Transconductance multipliers are employed to imple...

We propose a new architecture of an intelligent fuzzy control system. The basic idea of the system is based on analyzing error responses of a fuzzy control system to adjust its parameters properly. Two simulation results have shown that the proposed system is promising

A simplified structure of the ordinary delta-sigma decoder is described. The basic algorithm is derived and circuit diagrams for analogue and digital implementations are proposed

In this paper, we propose a new method to generalize Bayesian methods both for fuzzy likelihood and fuzzy prior probabilities. Based on interval Bayesian inference developed by Pan and Klir [1996], the proposed method overcomes the difficulty in developing a normalized fuzzy Bayesian inference recognized in the literature [Friihwirth-Schnatter, 199...

In this paper, we first distinguish two types of fuzzy controllers, Mamdani fuzzy controllers and fuzzy logical controllers. Mamdani fuzzy controllers are based on the idea of interpolation while fuzzy logical controllers are based on fuzzy logic in its narrow sense, i.e., fuzzy propositional logic. The two types of fuzzy controllers treat IF-THEN...

In this chapter, we investigate the following problem: given a data set involving n variables, determine key variables that contribute most to a specific partition of this data set. This problem has a broad applicability, even though it emerged in the context of a particular engineering application—the process of manufacturing electric circuit boar...

We propose fuzzy Bayesian inference on the basis of interval Bayesian inference developed by Pan and Klir. The proposed method overcomes the difficulty in developing a normalized fuzzy Bayesian inference recognized in the literature

We first prove that for a given set of data there exists a fuzzy measure fitting exactly the data if and only if there exists an exact solution of the associated fuzzy relation equation. Secondly, we continue to study the special neural network we proposed in Proc. IFSA'95 World Congress, pp. 61-64 (1995), and describe a learning algorithm for obta...

In this article, a new kind of reasoning for propositional knowledge, which is based on the fuzzy neural logic initialed by Teh, is introduced. A fundamental theorem is pre- sented showing that any fuzzy neural logic network can be represented by operations: bounded sum, complement, and scalar product. Propositional calculus of fuzzy neural logic i...

An abstract is not available.

A parametric approach based upon regular inference functions for approximate reasoning is presented. Inferences in approximate reasoning systems are realized by either the compositional rule of inference (CRI for short) or the compatibility modification inference (CMI for short). In general, CRI and conventional CMI inference methods do not satisfy...

A measure of nonspecificity is proposed for fuzzy sets with continuous membership functions defined, in general, on the n-dimensional Euclidean space R ⁿ. The proposed measure is restricted to fuzzy sets that are convex. The measure is justified in terms of the usual axiomatic requirements

The purpose of the paper is to show that, contrary to the claims of some probabilists of the Bayesian trade, probability theory relates to fuzzy set theory in a similar way as it relates to classical set theory. Examining the well known proof by R.T. Cox (1946), whose aim is to justify the rules of classical Bayesian inference, we point to some wea...

In this paper, a new approach to fuzzy clustering is introduced. This approach, which is based on the application of an evolutionary strategy to the fuzzy c-means clustering algorithm, utilizes the relationship between the various definitions of distance and structures implied in each given data set. As soon as a particular definition of distance i...

In this paper, normal form based interval-valued fuzzy sets, which is initiated by Turksen in 1986, are investigated. It is shown that DNF(•) ⊆ CNF(•) holds for a newly introduced family, called g-family, of conjugate pairs of t-norms and t-conorms. Interval-valued fuzzy set (IVFS) based approximate reasoning is also studied. The generalized compos...

It is well-known that the composition rule of inference which was introduced by Zadeh [1973], does not satisfy the inference rule, modus ponens, in general. It seems a disadvantage of applying it. However, in this paper, we show that we can make modus ponens satisfied by slightly modifying the original IF-THEN rules in the rule base. That is, we ca...

In this paper, we generalize the definition of Sugeno integrals by utilizing the so-called median operations, which are a special kind of aggregation operations. The generalized integrals, which we call median integrals, possess almost all common properties of Sugeno integrals. The properties of median integrals are investigated and their significa...

Fuzzy sets were introduced by Zadeh [9] in 1965 to represent/manipu-late data and information possessing nonstatistical uncertainties. Fuzzy sets serve as a means of representing and manipulating data that are not precise, but rather fuzzy.

Fuzzy logic methods have been used successfully in many real-world applications, but the foundations of fuzzy logic remain under attack. Taken together, these two facts constitute a paradox. A second paradox is that almost all of the successful fuzzy logic applications are embedded controllers, while most of the theoretical papers on fuzzy methods...

We generalize our previous results regarding approximate solutions of fuzzy relation equations (Yuan and Klir, 1993) to systems of fuzzy relation equations. By employing the equality index proposed by Gottwald (1985), we introduce a goodness measure of the performance of approximate solutions and derive a lower bound and an upper bound of solvabili...

Fuzzy logic methods have been used successfully in many real-world applications, but the foundations of fuzzy logic remain under attack. Taken together, these two facts constitute a paradox. A second paradox is that almost all of the successful fuzzy ...

In this paper, a more generalized version of fuzzy method-of-cases that was initiated by Ruan Da, E. E. Kerre, et al. in [1] is introduced. The validity of generalized fuzzy method-of-cases (GFMC for short) is investigated under various implication operators.

We investigate the relationship between fuzzy ideals and fuzzy congruences on a distributive lattice and obtain that the lattice of fuzzy ideals is isomorphic to the lattice of fuzzy congruence on a generalized Boolean algebra. Finally, consider the products of fuzzy ideals.

This paper proposes a new model for secure communication channels between two parties. The new model assumes that adversaries are storage space bounded, but not computationally bounded. At the initial phase of the secret communication, both parties exchange a large amount of random bits so that adversaries are not able to save them due to the stora...

This paper proposes a new model for secure communication channels between two parties. The new model assumes that adversaries are storage space bounded, but not computationally bounded. At the initial phase of the secret communication, both parties exchange a large amount of random bits so that adversaries are not able to save them due to the stora...

Thesis (Ph. D.)--State University of New York at Binghamton, Watson School of Engineering and Applied Science, 1997) Includes bibliographical references.