Input Pairs Research Articles

Routine uncertainty propagation for the marine carbon dioxide system

Pairs of marine carbonate system variables are often used to calculate others, but those results are seldom reported with estimates of uncertainties. Although the procedure to propagate these uncertainties is well known, it has not been offered in public packages that compute marine carbonate chemistry, fundamental tools that are relied on by the community. To remedy this shortcoming, four of these packages were expanded to calculate sensitivities of computed variables with respect to each input variable and to use those sensitivities along with user-specified estimates of input uncertainties (standard uncertainties) to propagate uncertainties of calculated variables (combined standard uncertainties). Sensitivities from these packages agree with one another and with analytical solutions to within 0.01%; similar agreement among packages was found for the combined standard uncertainties. One package was used to quantify how propagated uncertainties vary among computed variables, seawater conditions, and the chosen pair of carbonate system variables that is used as input. The relative contributions to propagated uncertainties from the standard uncertainties of the input pair of measurements and various other input data (equilibrium constants etc) were explored with a new type of diagram. These error-space diagrams illustrate that further improvement beyond today's state-of-the-art measurement uncertainties for the input pair would generally be ineffective at reducing the combined standard uncertainties because the contribution from the constants is larger. Likewise, using much more uncertain measurements of the input pair does not always substantially worsen combined standard uncertainty. The constants that contribute most to combined standard uncertainties are generally K1 and K2, as expected. Yet more of the propagated uncertainty in the computed saturation states of aragonite and calcite comes from their solubility products. Thus percent relative combined standard uncertainties for the saturation states are larger than for the carbonate ion concentration. Routine propagation of these uncertainties should become standard practice.

Bonferroni Mean With Weighted Interaction

Bonferroni mean aggregates the interaction between all pairs of inputs from some $n$ -dimensional input vector. Therefore, it is able to capture the dependency structure between the inputs. Weighted version of the Bonferroni mean then assumes that each input has a possibly different weight. Such an approach poses several constraints that decrease the modeling flexibility of the dependency structure on the input space. In this paper, we present an overview of the different approaches to weighted Bonferroni mean and introduce the Bonferroni mean with weighted interaction where interactions between inputs are weighted rather than the inputs themselves. We show several important properties of the Bonferroni mean with weighted interaction and its connection to concepts studied before. Finally, using the described approach we design a system for emergency management that is able to raise an alarm in the case where certain criteria are met.

Deep Localized Metric Learning

Metric learning has been widely used in many visual analysis applications, which learns new distance metrics to measure the similarities of samples effectively. Conventional metric learning methods learn a single linear Mahalanobis metric, yet such linear projections are not powerful enough to capture the nonlinear relationships. Recently, deep metric learning approaches, such as discriminative deep metric learning and deep transfer metric learning, have been introduced to fully exploit the nonlinearity of samples by learning hierarchical nonlinear transformations. However, these methods only learn holistic metrics over the input space and are limited for the heterogeneous data sets, where data varies locally. In this paper, we propose a deep localized metric learning approach for visual recognition by learning multiple fine-grained deep localized metrics. We first learn $K$ local subspaces and one holistic subspace with the K-auto-encoders-based clustering. Then, given an input pair, we compute its localized distance on each learned subspace and obtain the final distance representation. Finally, we train the entire neural networks to ensure the distances of positive pairs smaller than negative pairs by a large margin. Experimental results on three visual recognition applications, including face recognition, person re-identification, and scene recognition, show that our DLML outperforms most existing metric learning approaches.

Single-shot and lensless complex-amplitude imaging with incoherent light based on machine learning

We present a single-shot incoherent light imaging method for simultaneously observing both amplitude and phase without any imaging optics, based on machine learning. In the proposed method, an object with a complex-amplitude field is illuminated with incoherent light and is captured by an image sensor with or without a coded aperture. The complex-amplitude field of the object is reconstructed from a single captured image using a state-of-the-art deep convolutional neural network, which is trained with a large number of input and output pairs. In experimental demonstrations, the proposed method was verified with a handwritten character database, and the effect of a coded aperture printed on an overhead projector film in the reconstruction was examined. Our method has advantages over conventional wavefront sensing techniques using incoherent light, namely simplification of the optical hardware and improved measurement speed. This study shows the importance and practical impact of machine learning techniques in various fields of optical sensing.

Analysis of the dynamics of land use change and its prediction based on the integration of remotely sensed data and CA-Markov model, in the upstream Citarum Watershed, West Java, Indonesia

ABSTRACTIn this research, the integration of remotely sensed data and Cellular Automata-Markov model (CA-Markov) have been used to analyze the dynamics of land use change and its prediction for the next year. Training phase for the CA-Markov model has been created based on the input pair of land use, which is the result of land use mapping using Maximum Likelihood (ML) algorithm. Three-map comparison has been used to evaluate process accuracy assessment of the training phase for the CA-Markov model. Furthermore, the simulation phase for the CA-Markov model can be used to predict land use map for the next year. The analyze of the dynamics of land use change and its prediction during the period 1990 to 2050 can be obtained that the land serves as a water absorbent surfaces such as primary forest, secondary forest and the mixed garden area continued to decline. Meanwhile, on build land area that can lead to reduced surface water absorbing tends to increase from year to year. The results of this research can be used as input for the next research, which aims to determine the impact of land use changes in hydrological conditions against flooding in the research area.

A 12-bit 150-MS/s Sub-Radix-3 SAR ADC With Switching Miller Capacitance Reduction

This paper presents a 0.9-V 1.5-mW 150-MS/s 12-bit successive-approximation-register analog-to-digital converter (SAR ADC) in 40-nm CMOS, which achieves a synergistic integration of multi-bit per cycle (M-bit/cycle) and sub-radix techniques with the comprehensive enhancement of speed and error tolerance. The proposed sub-radix-3 architecture uses a merged digital-to-analog converter (DAC) with a two-input comparator instead of the conventional separated DACs (signal and reference DACs) with a four-input comparator to effectively improve the matching performance and the power efficiency. Alternative-reference-switching (ARS) DACs are developed to reduce the total DAC effective switching capacitance by 47.4%. Offset injection technique is implemented to calibrate the comparator offset without the speed and power penalty by eliminating the additional output loading of capacitor array or the extra input pair. The SAR logic of each conversion cycle is simplified into one single tri-latch current mode logic (CML) to directly control the DAC switches which effectively reduces the SAR delay by 2/3. The achieved peak signal to noise and distortion ratio (SNDR), spurious-free dynamic range (SFDR), and Walden FoM are 61.7 dB, 74.4 dB, and 10.3-fJ/conversion-step, respectively.

Energy and Entropy Measures of Fuzzy Relations for Data Analysis.

We present a new method for assessing the strength of fuzzy rules with respect to a dataset, based on the measures of the greatest energy and smallest entropy of a fuzzy relation. Considering a fuzzy automaton (relation), in which A is the input fuzzy set and B the output fuzzy set, the fuzzy relation R1 with greatest energy provides information about the greatest strength of the input-output, and the fuzzy relation R2 with the smallest entropy provides information about uncertainty of the input-output relationship. We consider a new index of the fuzziness of the input-output based on R1 and R2. In our method, this index is calculated for each pair of input and output fuzzy sets in a fuzzy rule. A threshold value is set in order to choose the most relevant fuzzy rules with respect to the data.

Compact Multi-Label Learning

Embedding methods have shown promising performance in multi-label prediction, as they can discover the dependency of labels. Most embedding methods cannot well align the input and output, which leads to degradation in prediction performance. Besides, they suffer from expensive prediction computational costs when applied to large-scale datasets. To address the above issues, this paper proposes a Co-Hashing (CoH) method by formulating multi-label learning from the perspective of cross-view learning. CoH first regards the input and output as two views, and then aims to learn a common latent hamming space, where input and output pairs are compressed into compact binary embeddings. CoH enjoys two key benefits: 1) the input and output can be well aligned, and their correlations are explored; 2) the prediction is very efficient using fast cross-view kNN search in the hamming space. Moreover, we provide the generalization error bound for our method. Extensive experiments on eight real-world datasets demonstrate the superiority of the proposed CoH over the state-of-the-art methods in terms of both prediction accuracy and efficiency.

Prediction of permanent deformation in asphalt pavements using a novel symbiotic organisms search–least squares support vector regression

The prediction of asphalt performance can be very important in terms of increasing service life and performance while saving energy and money. In this study, a new hybrid artificial intelligence (AI) system, SOS–LSSVR, has been proposed to predict the permanent deformation potential of asphalt pavement mixtures. SOS–LSSVR utilizes the symbiotic organisms search (SOS) and the least squares support vector regression (LSSVR), which are seen as a complementary system. The prediction model can be established from all input and output data pairs for LSSVR, while SOS optimizes the system’s tuning parameters. To avoid sampling bias and to partition the dataset into testing and training, a cross-validation technique was chosen. The results can be compared to those of previous studies and other predictive methods. Through the use of four error indicators, SOS–LSSVR accuracy was verified in predicting the permanent deformation behavior of an asphalt mixture. The present study demonstrates that the proposed AI system is a valuable decision-making tool for road designers. Additionally, the success of SOS–LSSVR in building an accurate prediction model suggests that the proposed self-optimized prediction framework has found an underlying pattern in the current database and thus can potentially be implemented in various disciplines.

Effects of phase correlations in naturalistic stimuli on quantitative information coding by fly photoreceptors.

Natural visual scenes are rarely random. Instead, intensity and wavelength change slowly in time and space over many regions of the scene, so that neighboring temporal and spatial visual inputs are more correlated and contain less information than truly random signals. It has been suggested that sensory optimization to match these higher order correlations (HOC) occurs at the earliest visual stages, and that photoreceptors can process temporal natural signals more efficiently than random signals. We tested this early-stage hypothesis by comparing the information content of Calliphora vicina photoreceptor responses to naturalistic inputs before and after removing HOC by randomizing phase. Forty different, 60-s long, naturalistic sequences (NS) were used, together with randomized-phase versions of the same sequences to give pink noise (PN) so that each input pair had identical means, variances, mean contrasts, and power spectra. We measured the information content of inputs and membrane potential responses by three different methods: coherence, mutual information, and compression entropy. We also used entropy and phase statistics of each pair as measures of HOC. Responses to randomized signals generally had higher gain, signal-to-noise ratio, and information rates than responses to NS. Information rate increased with a strong, positive, linear correlation to phase randomization within sequence pairs. This was confirmed by varying the degree of phase randomization. Our data indicate that individual photoreceptors encode input information by Weber's law, with HOC within natural sequences reducing information transfer by decreasing the number of local contrast events that exceed the noise-imposed threshold. NEW & NOTEWORTHY Natural visual scenes feature statistical regularities, or higher order correlations (HOC), both in time and space, to encode surfaces, textures, and object boundaries. Visual systems rely on this information; however, it remains controversial whether individual photoreceptors can discriminate and enhance information encoded in HOC. Here we show that the more HOC the stimulus contains, the lower the information transfer rate of photoreceptors. We explain our findings by applying the Weber's paradigm of differential signal perception.

A requiem for the Australian motor vehicle industry

ABSTRACTThe last automobiles manufactured in Australia rolled off the assembly line in the fall of 2017. This article looks at some of the factors that have impacted the industry since 1968 and led to its demise, including a high value of the Australian dollar in recent years, strategic decisions on the part of parent companies and reductions in governmental support and tariff protection. We estimate a cost function for the industry with inputs of domestic capital and labour and insourced intermediate goods as well as imported intermediate goods. The findings include that the remaining firms are operating in an output range of strongly statistically significant economies of scale, and that all of the input pairs are substitutes except for statistically significant complementary relationships between capital and domestic intermediate goods and labour and foreign intermediate goods. Unexpected results are that an increase in output per assembly plant appears to have a positive effect on total cost, while an increase in the effective tariff and an increase in the number of models appears to have a negative effect. One explanation for these robust but unexpected findings may be that total profit contribution is a part of total cost, and, therefore, factors that increase total profit contribution will also increase total cost.

Adaptive Signal Selection of Wide-Area Damping Controllers Under Various Operating Conditions

Since operating conditions of power systems always change, the input and output signals of wide-area damping controller (WADC), which are selected at an operating point, may not be able to guarantee the damping effect at other operating points. This paper focuses on a new adaptive signal selection for WADC against several operating conditions, such as various load demands, control signal failure, line and generator outages, and effect of communication latency. The joint controllability and observability is used to determine the best input and output pairs of WADC at any operating points. Small-signal and transient stabilities study in the IEEE 50-machine system including renewable sources, i.e., wind and solar photovoltaic generators are conducted to evaluate the effect of the proposed method. Study result demonstrates that the WADC with the adaptive signal selection yields superior damping effect to the WADC with the fixed signal selection over wide range operations.

Data Science Approaches for Electrochemical Engineers: An Introduction through Surrogate Model Development for Lithium-Ion Batteries

Data science, hailed as the fourth paradigm of science, is a rapidly growing field which has served to revolutionize the fields of bio-informatics and climate science and can provide significant speed improvements in the discovery of new materials, mechanisms, and simulations. Data science techniques are often used to analyze and predict experimental data, but they can also be used with simulated data to create surrogate models. Chief among the data science techniques in this application is machine learning (ML), which is an effective means for creating a predictive relationship between input and output vector pairs. Physics-based battery models, like the comprehensive pseudo-two-dimensional (P2D) model, offer increased physical insight, increased predictability, and an opportunity for optimization of battery performance which is not possible with equivalent circuit (EC) models. In this work, ML-based surrogate models are created and analyzed for accuracy and execution time. Decision trees (DTs), random forests (RFs), and gradient boosted machines (GBMs) are shown to offer trade-offs between training time, execution time, and accuracy. Their ability to predict the dynamic behavior of the physics-based model are examined and the corresponding execution times are extremely encouraging for use in time-critical applications while still maintaining very high (∼99%) accuracy.

The Effects of Comparator Dynamic Capacitor Mismatch in SAR ADC and Correction

This paper proposes a method of correcting the nonlinear parasitic capacitor of the input pair of comparator in successive approximations analog-to-digital converters (SAR ADCs). The correction method is proposed for the conventional binary-weighted capacitor array topology used in most of high resolution and high speed SAR ADCs. The effects of dynamic capacitor mismatch are first analyzed and then two feasible correction schemes are proposed to mitigate the impact of the nonlinear parasitic capacitor of the comparator. To verify the effectiveness of the proposed method, we designed a SAR ADC in a CMOS 40 nm process and characterized the design by intensive post-simulations. With the proposed correction schemes, the SFDR and SNDR of the SAR ADC increase about 7 and 4 dB, respectively, the differential nonlinearity and integral nonlinearity after calibration are improved from 1.00 and 3.81 to 0.67 LSB/0.57 LSB and 1.46 LSB/0.77 LSB, respectively.

Error Modeling And Compensation Of 3D Scanning Robot System Based On Pso-Rbfnn

abstract In order to improve the measurement accuracy of three-dimensional (3D) scanning robot, a method of 3D scanning robot system error modeling and compensation based on particle swarm optimization radial basis function neural network (PSO-RBFNN) is proposed to achieve intelligent compensation of measurement error. The structure, calibration and error modeling process of 3D scanning robot system are mainly described. Cleverly using the iterative closest point (ICP) algorithm to construct input and output data pairs of neural network, and the specific process of error modeling using PSO-RBFNN is given. Finally, through the actual experiment we test and verify the correctness and effectiveness of the proposed error modeling and compensation method by measuring the distance between the centers of two standard balls. Experimental results show: the proposed error model and the compensation method can effectively compensate the measurement errors and improve the accuracy of the 3D scanning robot system.

Fuzzy iterative PID-type control of vortex induced vibration suppression

This paper presents the application of Fuzzy Iterative PID-Type controller in suppressing the vortex induced vibration of a flexibly mounted cylindrical pipe. A pair of input and output data, which are the water flow speed and vibration amplitude, respectively have been measured through experimental work in generating a linear autoregressive with exogenous input, ARX model. Recursive least square was utilized in determining the appropriate parameters of stated model. It was found that the 2nd model order had fulfilled the selection requirements, which include lowest mean squared error, stable transfer function and biased correlation of error. Fuzzy Iterative PID and PD controllers have been developed and tested within MATLAB Simulink interface for the control purpose of the developed ARX model. It was found that the Fuzzy Iterative PD controller had remarkably attenuated the vibration up to 66.67 dB and tends to converge faster which gives good implication to the vibrating system.

Deep Learning: A Primer for Radiologists.

Deep learning is a class of machine learning methods that are gaining success and attracting interest in many domains, including computer vision, speech recognition, natural language processing, and playing games. Deep learning methods produce a mapping from raw inputs to desired outputs (eg, image classes). Unlike traditional machine learning methods, which require hand-engineered feature extraction from inputs, deep learning methods learn these features directly from data. With the advent of large datasets and increased computing power, these methods can produce models with exceptional performance. These models are multilayer artificial neural networks, loosely inspired by biologic neural systems. Weighted connections between nodes (neurons) in the network are iteratively adjusted based on example pairs of inputs and target outputs by back-propagating a corrective error signal through the network. For computer vision tasks, convolutional neural networks (CNNs) have proven to be effective. Recently, several clinical applications of CNNs have been proposed and studied in radiology for classification, detection, and segmentation tasks. This article reviews the key concepts of deep learning for clinical radiologists, discusses technical requirements, describes emerging applications in clinical radiology, and outlines limitations and future directions in this field. Radiologists should become familiar with the principles and potential applications of deep learning in medical imaging. ©RSNA, 2017.

High‐gain and power‐efficient dynamic amplifier for pipelined SAR ADCs

A new power-efficient dynamic residue amplifier for pipelined successive-approximation-register (SAR) ADCs is presented. A complementary input pair and a reference level detection technique are proposed to improve the performance of the dynamic amplifier. At a 400 MS/s sampling rate, the proposed amplifier achieves a gain of 16. The input referred noise is 57.2 μV and the total harmonic distortion is –57.1 dB. The power consumption of the proposed amplifier is 55.6 μW in 28 nm CMOS technology.

Radiation-Hardened CMOS Negative Voltage Reference for Aerospace Application

Voltage reference is the key module in analog and mixed-signal integrated circuits. This paper presents a radiation-hardened CMOS negative voltage reference for aerospace electronics. To improve the antiradiation performance, in the circuit design, the input pair of the operational amplifier is replaced from pMOS to nMOS. An extra unity-gain amplification stage is added and the compensation network is optimized. Besides, the start-up circuit is redesigned. In the layout design, the annular-gate structure is adopted to eliminate electric leakage, while the layout technique against single-event latch-up is also used. The prototype of the proposed circuit is fabricated using a bulk CMOS 0.6- $\mu \text{m}$ process with a $547\,\, \mu \text {m} \times 618\,\, \mu \text {m}$ chip area. By using the proposed circuit and layout optimizations, the measured temperature coefficient of reference is reduced to 13 ppm/°C and the output voltage drift is below 1.2% after 300-krad(Si) total ionizing dose. The measured single-event latch-up threshold is above 94.6 MeVcm2/mg.

A Coding Theorem for Bipartite Unitaries in Distributed Quantum Computation

We analyze implementations of bipartite unitaries by means of local operations and classical communication (LOCC) assisted by shared entanglement. We employ concepts and techniques developed in quantum Shannon theory to study an asymptotic scenario, in which two distant parties perform the same bipartite unitary on infinitely many pairs of inputs. We analyze minimum cost of entanglement and classical communication per copy. For two-round LOCC protocols, we derive a single-letter formula for the minimum cost of entanglement and classical communication, under an additional requirement that the error converges to zero faster than $1/n^4$, where $n$ is the number of input pairs. The formula is given by the "Markovianizing cost" of a tripartite state associated with the unitary, which can be computed by a finite-step algorithm. We also derive a lower bound on the minimum cost of resources, which applies for protocols with arbitrary number of rounds.