Structure-forming processes in Ca-induced whey protein isolate cold gelation

The structure-forming process of Ca²⁺-induced whey protein isolate (WPI) gels was studied, at 24°C, in the presence of 10 and 120 m CaCl₂. Pre-heating WPI suspensions (10% protein) at 90°C dramatically increased specific viscosity, but did not change the number of accessible sulfhydryl groups, compared to pre-heating at 70°C. The most important factor governing the process appeared to be CaCl₂ concentration, rather than the reactive sulfhydryl groups. At 10 m CaCl₂, the increase in aggregate size and network connectivity over time was achieved by clustering of adjacent aggregates. At 120 m CaCl₂, the increase in aggregate size and connectivity was by enlargement of the aggregates which formed connected paths and filled up interstitial spaces.     

Stability Constants of Oximes of 4-Substituted-2-acetylphenols [1]

Proton-ligand stability constants (pKd) and metal-ligand stability constants for six oximes of 4-substituted-2-acetyl phenol in H₂O: Dioxane (1:3) are obtained by the spectrophotometric method of Magnusson et al. as modified by Patel et al.     

A notion of task relatedness yielding provable multiple-task learning guarantees

The approach of learning multiple “related” tasks simultaneously has proven quite successful in practice; however, theoretical justification for this success has remained elusive. The starting point for previous work on multiple task learning has been that the tasks to be learned jointly are somehow “algorithmically related”, in the sense that the results of applying a specific learning algorithm to these tasks are assumed to be similar. We offer an alternative approach, defining relatedness of tasks on the basis of similarity between the example generating distributions that underlie these tasks. We provide a formal framework for this notion of task relatedness, which captures a sub-domain of the wide scope of issues in which one may apply a multiple task learning approach. Our notion of task similarity is relevant to a variety of real life multitask learning scenarios and allows the formal derivation of generalization bounds that are strictly stronger than the previously known bounds for both the learning-to-learn and the multitask learning scenarios. We give precise conditions under which our bounds guarantee generalization on the basis of smaller sample sizes than the standard single-task approach. Editors: Daniel Silver, Kristin Bennett, Richard Caruana. A preliminary version of this paper appears in the proceedings of COLT’03, (Ben-David and Schuller 2003).     

Using Bayesian networks and virtual coverage to hit hard-to-reach events

Reaching hard-to-reach coverage events is a difficult task that requires both time and expertise. Data-driven coverage directed generation (CDG) can assist in the task when the coverage events are part of a structured coverage model, but is a priori less useful when the target events are singular and not part of a model. We present a data-driven CDG technique based on Bayesian networks that can improve the coverage of cross-product coverage models. To improve the capability of the system, we also present virtual coverage models as a means for enabling data-driven CDG to reach singular events. A virtual coverage model is a structured coverage model (e.g., cross-product coverage) defined around the target event, such that the target event is a point in the structured model. The CDG system can exploit this structure to learn how to reach the target event from covered points in the structured model. A case study using CDG and virtual coverage to reach a hard-to-reach event in a multi-processor system demonstrates the usefulness of the proposed method.     

A theory of learning from different domains

Discriminative learning methods for classification perform well when training and test data are drawn from the same distribution. Often, however, we have plentiful labeled training data from a source domain but wish to learn a classifier which performs well on a target domain with a different distribution and little or no labeled training data. In this work we investigate two questions. First, under what conditions can a classifier trained from source data be expected to perform well on target data? Second, given a small amount of labeled target data, how should we combine it during training with the large amount of labeled source data to achieve the lowest target error at test time? We address the first question by bounding a classifier’s target error in terms of its source error and the divergence between the two domains. We give a classifier-induced divergence measure that can be estimated from finite, unlabeled samples from the domains. Under the assumption that there exists some hypothesis that performs well in both domains, we show that this quantity together with the empirical source error characterize the target error of a source-trained classifier. We answer the second question by bounding the target error of a model which minimizes a convex combination of the empirical source and target errors. Previous theoretical work has considered minimizing just the source error, just the target error, or weighting instances from the two domains equally. We show how to choose the optimal combination of source and target error as a function of the divergence, the sample sizes of both domains, and the complexity of the hypothesis class. The resulting bound generalizes the previously studied cases and is always at least as tight as a bound which considers minimizing only the target error or an equal weighting of source and target errors.     

Genetic-fuzzy approach to the Boolean satisfiability problem

This study is concerned with the Boolean satisfiability (SAT) problem and its solution in setting a hybrid computational intelligence environment of genetic and fuzzy computing. In this framework, fuzzy sets realize an embedding principle meaning that original two-valued (Boolean) functions under investigation are extended to their continuous counterparts resulting in the form of fuzzy (multivalued) functions. In the sequel, the SAT problem is reformulated for the fuzzy functions and solved using a genetic algorithm (GA). It is shown that a GA, especially its recursive version, is an efficient tool for handling multivariable SAT problems. Thorough experiments revealed that the recursive version of the GA can solve SAT problems with more than 1000 variables     

雷达回波数据实时采集与分析可视化软件开发

文中设计了雷达回波数据实时采集与分析的可视化软件,采用PCI50612数据采集卡进行二次开发,利用VC 编程技术以时域和频域图的方式实时显示模拟雷达回波以及分析结果,并保存数据供后期分析使用.具有连接方便、操作简单、高速大容量雷达数据采集等优点,对雷达回波数据的后期分析应用具有指导意义.     

Efficient Commitment Schemes with Bounded Sender and Unbounded Receiver

In this paper we address the problem of constructing commitment schemes where the sender is bounded to polynomial time and the receiver may be all powerful. Many known constructions for such commitment schemes are based on the hardness of factoring large integers. However, these schemes typically use integers of a special form and thus require a rather expensive initialization procedure for establishing these special-form integers. In this paper we present a scheme which is based on the hardness of factoring large integers but avoids the need of a complex initialization procedure. Received 31 May 1996 and revised 26 December 1996     

水下无线传感器网络中一种安全定位算法

提出一种基于阵列传输结构的无线传感器网络安全定位算法（USA）。该算法主要解决水下无线传感器网络（UWSN）面临的一些安全威胁问题。以提高无线传感器网络安全性,特别是位置信息的安全性为设计目标。利用节点协作形成的阵列作为天线阵列进行相互通信,在不增加额外硬件成本的同时,还获得阵列天线给无线传感器网络带来的优势,如减小多径效应、提高接收端的信噪比、增加系统容量等。USA算法基于这种阵列结构使网络得到很高安全特性,特别是,对Wormhole攻击具有非常好的抵御性能。仿真实验证明该算法的有效性。