A 20-MHz sixth-order BiCMOS parasitic-insensitive continuous-timefilter and second-order equalizer optimized for disk-drive read channels

A fast parasitic-insensitive continuous-time filter and equalizer integrated circuit that uses an active integrator is described. Circuit techniques for excess-phase cancellation and for setting the corner-frequency of the filter and equalizer are also described. These techniques result in a filter and equalizer chip with performance independent of process, supply, and temperature without employing phase-lock loops. The 20-MHz sixth-order Bessel filter and second-order equalizer operate from 5 V and generate only 0.24% (-52 dB) of total harmonic distortion when processing 2-V_pp differential output signals. The device is optimized to limit high-frequency noise and to amplitude equalize the data pulses in hard disk read-channel systems. The device supports data rates of up to 36 Mb/s and is built in a 1.5-μm, 4-GHz BiCMOS technology     

A positive-feedback transconductance amplifier with applications tohigh-frequency, high-Q CMOS switched-capacitor filters

The authors describe a high-bandwidth amplifier that simultaneously achieves high gain by using internal positive feedback. The results obtained when the amplifier is used in a general-purpose switched-capacitor biquadratic building block are also presented. This device achieves 150-kHz center-frequency operation with Q accuracy of 15% when clocked at 7.5 MHz with 10-V ±10% supplies from -55°C to +125°C. With a minimum power supply of 4.5 V, this filter operates with center frequencies up to 100 kHz when clocked at 5 MHz, while performing with the same accuracy and across the same temperature range as above     

Wear Resistance of Coatings from Self‐Fluxing Alloys after Laser Doping under Dry Friction Conditions

The wear resistance of Ni–Cr–B–Si–C coatings after laser treatment with a doping covering under dry friction conditions has been investigated. The microstructure of coatings after fusion by a gas burner and a laser beam and the roughness and waviness parameters of the friction surfaces before and after wear have been investigated.     

Bounding the Inefficiency of Length-Restricted Prefix Codes

Abstract. We consider an alphabet Σ= {a ₁ ,\ldots,a _n } with corresponding symbol probabilities p ₁ ,\ldots,p _n . For each prefix code associated to Σ , let l _i be the length of the codeword associated to a _i . The average code length c is defined by c=\sum _i=1 ⁿ p _i l _i . An optimal prefix code for Σ is one that minimizes c . An optimal L -restricted prefix code is a prefix code that minimizes c constrained to l _i ≤ L for i=1,\ldots,n . The value of the length restriction L is an integer no smaller than \lceil log n \rceil . Let A be the average length of an optimal prefix code for Σ . Also let A _L be the average length of an optimal L -restricted prefix code for Σ . The average code length difference ɛ is defined by ɛ=A _L -A . Let ψ be the golden ratio 1.618. In this paper we show that ɛ < 1/ψ ^{L-\lceil\log (n+\lceil\log n\rceil-L)\rceil-1} when L > \lceil log n \rceil . We also prove the sharp bound ɛ < \lceil log n \rceil -1 , when L = \lceil log n \rceil . By showing the lower bound 1/(ψ ^{L-\lceil\log n\rceil+2+\lceil\log (n/(n-L))\rceil} -1) on the maximum value of ɛ , we guarantee that our bound is asymptotically tight in the range \lceil log n \rceil < L ≤ n/2 . When L\geq \lceil log n \rceil +11 , the bound guarantees that ɛ < 0.01 . From a practical point of view, this is a negligible loss of compression efficiency. Furthermore, we present an O(n) time and space 1/ψ ^{L-\lceil\log (n+\lceil\log n\rceil-L)\rceil-1} -approximative algorithm to construct L -restricted prefix codes, assuming that the given probabilities are already sorted. The results presented in this paper suggest that one can efficiently implement length restricted prefix codes, obtaining also very effective codes.     

Improved Approximation Algorithms for the Average-Case Tree Searching Problem

We study the following tree search problem: in a given tree T=(V,E) a vertex has been marked and we want to identify it. In order to locate the marked vertex, we can use edge queries. An edge query e asks in which of the two connected components of T?e the marked vertex lies. The worst-case scenario where one is interested in minimizing the maximum number of queries is well understood, and linear time algorithms are known for finding an optimal search strategy. Here we study the more involved average-case analysis: A function w:V→?⁺ is given which measures the likelihood for a vertex to be the one marked, and we seek to determine the strategy (decision tree) that minimizes the weighted average number of queries. In a companion paper we prove that the above tree search problem is $\mathcal {NP}$ -complete even for the class of trees of bounded diameter or bounded degree. Here, we match this complexity result with a tight algorithmic analysis of the bounded degree instances. We show that any optimal strategy (i.e., one that minimizes the weighted average number of queries) performs at most O(Δ(T)(log|V|+log(w(T)/w _min))) queries in the worst case, where w(T) is the sum of the likelihoods of the vertices of T, w _min is the minimum positive likelihood over the vertices of T and Δ(T) is the maximum degree of T. We combine this result with a non-trivial exponential time algorithm to provide an FPTAS for trees with bounded degree. We also show that for unbounded instances a natural greedy strategy attains a 1.62-approximation, improving upon the best known 14-approximation guarantee, previously provided by two of the authors.     

Comparison of microscopic and flow cytometric detection of platelet antibody in dogs suspected of having immune-mediated thrombocytopenia

A flow cytometric platelet immunofluorescence assay (FC-PIFA) was compared with a previously developed microscopic platelet immunofluorescence assay (MI-PIFA) for detection of circulating platelet antibody. Both assays were performed on serum from 10 healthy dogs with normal platelet count, and on serum from 27 thrombocytopenic dogs--18 had primary immune-mediated thrombocytopenia (IMT), and 9 had IMT in addition to other immune-mediated disease (secondary IMT). Both assays yielded negative results for all control dogs. The MI-PIFA and FC-PIFA results were in agreement in 23 dogs with IMT (14 positive and 9 negative). There was linear correlation between MI-PIFA scores and FC-PIFA results (r = 0.873). Positive results were obtained for 55.5% of the dogs with suspected IMT, using the MI-PIFA, compared with 67%, using the FC-PIFA; however, the difference was not statistically significant. Use of fresh or frozen fixed donor platelets as the antigen source yielded similar results in the FC-PIFA.     

A BiCMOS continuous-time filter for video signal processingapplications

A fully integrated BiCMOS continuous-time filter for video signal processing applications is presented. It incorporates an input clamping circuit, a third-order equalizer, a fifth-order elliptic filter with sinx/x correction, and a 75-Ω driver. The architectures of the input and output amplifiers as well as the filter and the equalizer are chosen based on the extensive study of circuit structures and Monte Carlo simulation to meet the linearity requirement for the broadcast-quality video system. The complete chip achieves a low-pass filter response with a 5.5-MHz cutoff frequency (f_cc), 0.3-dB passband ripple, 20-ns group delay variation up to 0.9 f_c, and 43-dB attenuation at 1.45 f_c. With a nominal 2-Vpp signal at the output, measured results show 0.2% differential gain, 0.38° differential phase, and 1.7-mV rms noise demonstrating 10-bit linearity in a 1.5-μm 4-GHz BiCMOS process technology. The filter active area is 8 mm² and it dissipates 350 mW in a single 5-V power supply     

Informing sequential clinical decision-making through?reinforcement learning: an empirical study

This paper highlights the role that reinforcement learning can play in the optimization of treatment policies for chronic illnesses. Before applying any off-the-shelf reinforcement learning methods in this setting, we must first tackle a number of challenges. We outline some of these challenges and present methods for overcoming them. First, we describe a multiple imputation approach to overcome the problem of missing data. Second, we discuss the use of function approximation in the context of a highly variable observation set. Finally, we discuss approaches to summarizing the evidence in the data for recommending a particular action and quantifying the uncertainty around the Q-function of the recommended policy. We present the results of applying these methods to real clinical trial data of patients with schizophrenia.     

Three space-economical algorithms for calculatingminimum-redundancy prefix codes

The minimum-redundancy prefix code problem is to determine, for a given list W=[ω₁,..., ω_n] of n positive symbol weights, a list L=[l₁,...,l_n] of n corresponding integer codeword lengths such that Σ_i=1 ⁿ 2^-li⩽1 and Σ_i=1ⁿ ω_il_i is minimized. Let us consider the case where W is already sorted. In this case, the output list L can be represented by a list M=[m₁,..., m_H], where m_l, for l=1,...,H, denotes the multiplicity of the codeword length l in L and H is the length of the greatest codeword. Fortunately, H is proved to be O(min(log(1/p₁),n)), where p₁ is the smallest symbol probability, given by ω₁/Σ _i=1ⁿ ω_i. We present the Fast LazyHuff (F-LazyHuff), the Economical LazyHuff (E-LazyHuff), and the Best LazyHuff (B-LazyHuff) algorithms. F-LazyHuff runs in O(n) time but requires O(min(H², n)) additional space. On the other hand, E-LazyHuff runs in O(n+nlog(n/H)) time, requiring only O(H) additional space. Finally, B-LazyHuff asymptotically overcomes, the previous algorithms, requiring only O(n) time and O(H) additional space. Moreover, our three algorithms have the advantage of not writing over the input buffer during code calculation, a feature that is very useful in some applications     

Assessing Tuning Parameter Selection Variability in Penalized Regression

Penalized regression methods that perform simultaneous model selection and estimation are ubiquitous in statistical modeling. The use of such methods is often unavoidable as manual inspection of all possible models quickly becomes intractable when there are more than a handful of predictors. However, automated methods usually fail to incorporate domain-knowledge, exploratory analyses, or other factors that might guide a more interactive model-building approach. A hybrid approach is to use penalized regression to identify a set of candidate models and then to use interactive model-building to examine this candidate set more closely. To identify a set of candidate models, we derive point and interval estimators of the probability that each model along a solution path will minimize a given model selection criterion, for example, Akaike information criterion, Bayesian information criterion (AIC, BIC), etc., conditional on the observed solution path. Then models with a high probability of selection are considered for further examination. Thus, the proposed methodology attempts to strike a balance between algorithmic modeling approaches that are computationally efficient but fail to incorporate expert knowledge, and interactive modeling approaches that are labor intensive but informed by experience, intuition, and domain knowledge. Supplementary materials for this article are available online.