Stevens Institute of Technology

About: Stevens Institute of Technology is a education organization based out in Hoboken, New Jersey, United States. It is known for research contribution in the topics: Cognitive radio & Wireless network. The organization has 5440 authors who have published 12684 publications receiving 296875 citations. The organization is also known as: Stevens & Stevens Tech.

A Survey for Cervical Cytopathology Image Analysis Using Deep Learning

Abstract: Cervical cancer is one of the most common and deadliest cancers among women. Despite that, this cancer is entirely treatable if it is detected at a precancerous stage. Pap smear test is the most extensively performed screening method for early detection of cervical cancer. However, this hand-operated screening approach suffers from a high false-positive result because of human errors. To improve the accuracy and manual screening practice, computer-aided diagnosis methods based on deep learning is developed widely to segment and classify the cervical cytology images automatically. In this survey, we provide a comprehensive study of the state of the art approaches based on deep learning for the analysis of cervical cytology images. Firstly, we introduce deep learning and its simplified architectures that have been used in this field. Secondly, we discuss the publicly available cervical cytopathology datasets and evaluation metrics for segmentation and classification tasks. Then, a thorough review of the recent development of deep learning for the segmentation and classification of cervical cytology images is presented. Finally, we investigate the existing methodology along with the most suitable techniques for the analysis of pap smear cells.

Portfolio optimization with drawdown constraints

Abstract: We propose a new one-parameter family of risk functions defined on portfolio return sample -paths, which is called conditional drawdown-at-risk (CDaR). These risk functions depend on the portfolio drawdown (underwater) curve considered in active portfolio management. For some value of the tolerance parameter α , the CDaR is defined as the mean of the worst

Adaptive downlink scheduling and rate selection: a cross-layer design

Abstract: In this paper, we discuss a cross-layer design for joint user scheduling and adaptive rate control for downlink wireless transmission. We take a stochastic learning-based approach to achieve this. The scheduling is performed at the medium access control (MAC) layer, whereas the rate selection takes place at the physical/link (PHY/LINK) layer. These two components residing in the two layers exchange information to ensure that user defined rate requests are satisfied by the right combination of transmission schedules and rate selections. The method is highly efficient for low mobility applications with mobile speeds in the order of a few kilometers per hour. While simple to implement, this technique requires no explicit channel estimation phase. The only feedback used are the single bit ACK/NACK signal indicating the correct reception/failure of the packet. As shown in the convergence theorems, the algorithm achieves optimal performance in "stationary" channels. With slowly varying channels, the rate selection algorithm sees a "quasi-stationary" channel and adaptively converges to an optimal solution. Simulations performed using a third-generation wireless system, namely, high-speed downlink packet access (HSDPA) validate the theoretical results.

Active Control of Translating Media With Arbitrarily Varying Length

Abstract: The general control laws for pointwise controllers to dissipate vibratory energies of translating beams and strings with arbitrarily varying length are presented. Special domain and boundary control laws that can be easily implemented result as a special case. Sufficient conditions for uniform stability and uniform exponential stability of controlled systems are established via Lyapunov stability criteria. Numerical simulations demonstrate the effectiveness of the active controllers in stabilizing translating media during both extension and retraction. Optimal gains leading to the fastest rates of decay of vibratory energies of controlled systems are identified. It is shown that under the optimal control gains, translating media can be completely stabilized during extension and retraction.

On the homeotopy group of a non-orientable surface

Abstract: Let X be a closed, compact connected 2-manifold (a surface ), which we will denote by O or N if we wish to stress that X is orientable or non-orientable. Let G ( X ) denote the group of all homeomorphisms X → X , D ( X ) the normal subgroup of homeomorphisms isotopic to the identity, and H ( X ) the factor group G ( X )/ D ( X ), i.e. the homeotopy group of X . The problem of determining generators for H ( O ) was considered by Lickorish in ( 7, 8 ), and the second of these papers specifies a finite set of generators of a particularly simple type. In (10) and (11) Lickorish considered the analogous problem for non-orientable surfaces, and, using Lickorish's partial results, Chilling-worth (4) determined a finite set of generators for H ( N ). While the generators obtained for H ( O ) and H ( N ) were strikingly similar, it was noteworthy that the techniques used in the two cases were different, and in particular that little use was made in the non-orientable case of the earlier results obtained on the orientable case. The purpose of this note is to show that the results of Lickorish and Chillingworth on non-orientable surfaces follow rather easily from the work in ( 7, 8 ) by an application of some ideas from the theory of covering spaces (2). Moreover, while Lickorish and Chillingworth sought only to find generators , we are able to show (Theorem 1) how in fact the entire structure of the group H ( N ) is determined by H ( O ), where O is an orientable double cover of N. Finally, we are able to determine defining relations for H ( N ) for the case where N is the connected sum of 3 projective planes (Theorem 3).