学位级别： M.S.C.S., Master of Science in Computer Science/ Master of Science in Clinical Science
Over the years, machine learning techniques have been used in a wide variety of security sensitive applications due to the high reliability and accuracy of its results. But recent findings in the domain of adversarial...
详细信息
Over the years, machine learning techniques have been used in a wide variety of security sensitive applications due to the high reliability and accuracy of its results. But recent findings in the domain of adversarial machine learning have shown that such deep learning models could be potentially vulnerable to attacks. A backdoor attack is one such attack where malicious data containing a predefined perturbation is added to the training data so that when the model is trained on it, a backdoor is created. This backdoor is generally hidden and can only be activated when the attacker adds the perturbation to the test data. In the domain of natural language processing, such poisoned data is generally created by adding a sequence of trigger words and changing the label of the data to the target class. But these attacks can be easily detected by visual inspection since the context of the poisoned text does not resemble its label. That is why to hide the poisoned data better, we have come up with a novel approach to generate poisoned data that modifies the text in such a way that the label fits the context of the poisoned text. Our attack algorithm called SentMod can achieve an attack success ratio of 97% by poisoning only 2% of the training data. We run extensive experiments on multiple deep learning models using different datasets to verify the effectiveness of our attack method.
In this thesis I explore the place of the actualized utopian American alternative community of higher learning within a philosophy of human culture. I carry out such a study in order to articulate a unifying descripti...
详细信息
In this thesis I explore the place of the actualized utopian American alternative community of higher learning within a philosophy of human culture. I carry out such a study in order to articulate a unifying description of movement; to find the “target” that orients the activity of actualized utopia, as opposed to a classificatory rule to define a mechanically determined utopian-object. Although these communities have varied in practice since the first generation of them were created in the mid-nineteenth century, I argue that they have played, and continue to play, a particular role in America. They imagine what a world beyond their own could look like; and through their resistance to the perceived status quo (on which they rely for their identity), they prefigure a world that could be. Although this world never comes fully into existence, the actualization (i.e., the process of becoming a dynamic symbolic product of culture) of these communities illuminates new possibilities for human life in the larger culture. While these actualized utopian communities often live long beyond their pre-figurative years, paradoxically their practices largely stay the same. A resistance to the perceived modern culture, and a vision of a world that offers more opportunities for life and dignity, has made these communities places that cultivate persons with a sense of agency. These communities create a sense of active consciousness; they create people who believe the world is changeable. An open humanistic and personalistic theodicy is what distinguishes these schools from a cult of domination, or an organization bent on a utopia that depersonalizes and dehumanizes the opportunities of others in service of a singular vision.
Advisors/Committee Members: Auxier, Randall
In the present thesis, we deal with the construction of non trivial formulas in higherorder logic languages. In particular, we focus on using SO (Second-Order logic) andTO (Third-Order logic) to express SATQBFk, and S...
详细信息
In the present thesis, we deal with the construction of non trivial formulas in higherorder logic languages. In particular, we focus on using SO (Second-Order logic) andTO (Third-Order logic) to express SATQBFk, and SATQBF respectively. First of all,we explain the relationship between logic languages and complexity classes. Then wegive formal definitions and examples for FO (First-Order), SO andHO (i≥2) (Higher-Order logic). It is known that, for every k ≥1, SATQBFk is acomplete problem for the level Σ [superscript P subscript k] of PH (Polynomial-time hierarchy), andthat SATQBF is a complete problem for PSPACE. As the expressibility of SO isknown to equal the class PH, then we know that there must be an SO formula whichcan express SATQBFk. On the other hand, PSPACE is known to be equal inexpressive power to SO with the addition of a second order transitive closurequantifier, which is widely conjectured to be strictly more expressive than SO *** TO includes PSPACE , this means that there must be a TO formula that canexpress SATQBF . Here we give first a top down explanation on the use of SO toexpress SATQBFk. A detailed SO formula is presented. We then give a top downpresentation of the sketch of a TO formula for SATQBF
Multiple description coding (MDC) using Compressive Sensing (CS) mainly aims at restoring an image from a small subset of samples with reasonable accuracy using an iterative message passing decoding algorithm commonly...
详细信息
Multiple description coding (MDC) using Compressive Sensing (CS) mainly aims at restoring an image from a small subset of samples with reasonable accuracy using an iterative message passing decoding algorithm commonly known as Belief Propagation (BP). The CS technique can accurately recover any compressible or sparse signal from a lesser number of non-adaptive, randomized linear projection samples than that specified by the Nyquist *** this work, we demonstrate how CS-based encoding generates measurements from the sparse image signal and the measurement matrix. Then we demonstrate how a BP decoding algorithm reconstructs the image from the measurements generated. In our work, the CS-BP algorithm assumes that all the unknown variables have the same prior distribution as we do not have any knowledge of the side information available during the initiation of the decoding process. Thus, we prove that this algorithm is effective even in the absence of side information
Today, many resources are freely available on the Internet in the form of PDF documents. However, free PDF documents may not contain what people expect. There are several ways that allow attackers/viruses (e.g., Code ...
详细信息
Today, many resources are freely available on the Internet in the form of PDF documents. However, free PDF documents may not contain what people expect. There are several ways that allow attackers/viruses (e.g., Code Red, Melissa) to add malicious content to PDF files, which can badly harm your devices (e.g., redirect you to a fake website, corrupt your operating system, get full access to your devices, etc.). This project aims to detect potential malicious content in PDF files. There are several types of malicious content, such as executable JavaScript, shellcode, and adware. We will first collect the PDF file details and save them to a csv file. We will use the csv file data and test it with our machine learning model, which we build using the Random Forest (RF) algorithm. A result of prediction will be given to a Chrome browser extension. It will run a detector in the background to detect malicious JavaScript from the PDF file and notify users whether the PDF file is benign or malicious during downloading.
Data Mining algorithms is a practical, technically-oriented guide to data mining algorithms that covers the most important algorithms for building classification, regression, and clustering models, as well as techniqu...
详细信息
ISBN:
(数字)9781118950807
ISBN:
(纸本)9781118950951
Data Mining algorithms is a practical, technically-oriented guide to data mining algorithms that covers the most important algorithms for building classification, regression, and clustering models, as well as techniques used for attribute selection and transformation, model quality evaluation, and creating model ensembles. The author presents many of the important topics and methodologies widely used in data mining, whilst demonstrating the internal operation and usage of data mining algorithms using examples in R.
Industry has been heavily pushing for new methods of human-computer interactionin the last several years and this has seen many different technologiesmove into the mainstream, from infrared sensors[1] and z-cameras[2]...
详细信息
Industry has been heavily pushing for new methods of human-computer interactionin the last several years and this has seen many different technologiesmove into the mainstream, from infrared sensors[1] and z-cameras[2] to touchscreens[3]. Because touch screens are a more mature and developed technologythey provide an ideal platform for mainstream technology development, but theirlevel of interactiveness is limited, and these limitations must be overcome or compensatedfor with clever interface *** this thesis, a solution to these inherent limitations in touch screen interfacedesign is proposed by augmenting the touch screens interaction capabilities withone or two cameras to enable a near touch user experience on top of the standardtouch screen. This offers flexibility in system design (the near touch isimplemented as an extra layer, and can be activated only if present) as well asproviding an inexpensive solution. Several Image processing algorithms relevantto this task are also discussed and their implementation evaluated
暂无评论