Global optimization of case-based reasoning for breast cytology diagnosis

Case-based reasoning (CBR) is one of the most popular prediction techniques in medical domains because it is easy to apply, has no possibility of overfitting, and provides a good explanation for the output. However, it has a critical limitation – its prediction performance is generally lower than other AI techniques like artificial neural networks (ANN). In order to obtain accurate results from CBR, effective retrieval and matching of useful prior cases for the problem is essential, but it is still a controversial issue to design a good matching and retrieval mechanism for CBR systems. In this study, we propose a novel approach to enhance the prediction performance of CBR. Our suggestion is the simultaneous optimization of feature weights, instance selection, and the number of neighbors that combine using genetic algorithms (GA). Our model improves the prediction performance in three ways – (1) measuring similarity between cases more accurately by considering relative importance of each feature, (2) eliminating useless or erroneous reference cases, and (3) combining several similar cases represent significant patterns. To validate the usefulness of our model, this study applied it to a real-world case for evaluating cytological features derived directly from a digital scan of breast fine needle aspirate (FNA) slides. Experimental results showed that the prediction accuracy of conventional CBR may be improved significantly by using our model. We also found that our proposed model outperformed all the other optimized models for CBR using GA.     

Team THOR's Entry in the DARPA Robotics Challenge Finals 2015

This paper describes Team THOR's approach to human‐in‐the‐loop disaster response robotics for the 2015 DARPA Robotics Challenge (DRC) Finals. Under the duress of unpredictable networking and terrain, fluid operator interactions and dynamic disturbance rejection become major concerns for effective teleoperation. We present a humanoid robot designed to effectively traverse a disaster environment while allowing for a wide range of manipulation abilities. To complement the robot hardware, a hierarchical software foundation implements network strategies that provide real‐time feedback to an operator under restricted bandwidth using layered user interfaces. Our strategy for humanoid locomotion includes a backward‐facing knee configuration paired with specialized toe and heel lifting strategies that allow the robot to traverse difficult surfaces while rejecting external perturbations. With an upper body planner that encodes operator preferences, predictable motion plans are executed in unforeseen circumstances. These plans are critical for manipulation in unknown environments. Our approach was validated during the DRC Finals competition, where Team THOR scored three points in 18 min of operation time, and the results are presented along with an analysis of each task.     

Distributed filtering and synchronization of diffusively state‐coupled heterogeneous systems

This paper studies distributed filtering‐based ssynchronization of diffusively state‐coupled heterogeneous systems. For given heterogeneous subsystems and a network topology, sufficient conditions for the filtering‐based synchronization are developed with a guaranteed performance. The estimation and synchronization error dynamics are obtained in a decoupled form, and it is shown that the filter and the controller can be designed separately by LMIs. The feasibility of the proposed design method using LMIs is discussed, and the main results are validated through examples with various setup. Copyright © 2016 John Wiley & Sons, Ltd.     

R-D optimized progressive compression of 3D meshes using prioritized gate selection and curvature prediction

A rate-distortion (R-D) optimized progressive coding algorithm for three-dimensional (3D) meshes is proposed in this work. We propose the prioritized gate selection and the curvature prediction to improve the connectivity and geometry compression performance, respectively. Furthermore, based on the bit plane coding, we develop a progressive transmission method, which improves the qualities of intermediate meshes as well as that of the fully reconstructed mesh, and extend it to the view-dependent transmission method. Experiments on various 3D mesh models show that the proposed algorithm provides significantly better compression performance than the conventional algorithms, while supporting progressive reconstruction efficiently.     

Sliding mode-like fuzzy logic control with self-tuning the deadzone parameters

In this paper, a fuzzy logic controller (FLC) is designed based on the similarity between the FLC and the sliding mode control (SMC). The proposed scheme provides the sliding mode-like FLC with fast self-tuning the dead-zone parameters (boundary layer thickness) under parameter variations of the controlled system. To show the validity and the effectiveness of the proposed control method, simulations are performed for the position control of a rotary inverted pendulum     

Suboptimal filter for continuous‐time linear systems with unknown parameters

The filtering problem for continuous‐time linear systems with unknown parameters is considered. A new suboptimal filter is herein proposed. It is based on the optimal mean‐square linear combination of the local Kalman filters. In contrast to the optimal weights, the suboptimal weights do not depend on current observations; thus, the proposed filter can easily be implemented in real‐time. Examples demonstrate high accuracy and efficiency of the suboptimal filter. Copyright © 2008 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

Self-absorption correction for solid-state photoluminescence quantum yields obtained from integrating sphere measurements

A new method is presented for analyzing the effects of self-absorption on photoluminescence integrating sphere quantum yield measurements. Both the observed quantum yield and luminescence spectrum are used to determine the self-absorption probability, taking into account both the initial emission and subsequent absorption and reemission processes. The analysis is experimentally validated using the model system of the laser dye perylene red dispersed in a polymer film. This approach represents an improvement over previous methods that tend to overestimate the true quantum yield, especially in cases with high sample absorbance or quantum yield values.     

Simultaneous optimization of artificial neural networks for financial forecasting

Artificial neural networks (ANNs) have been popularly applied for stock market prediction, since they offer superlative learning ability. However, they often result in inconsistent and unpredictable performance in the prediction of noisy financial data due to the problems of determining factors involved in design. Prior studies have suggested genetic algorithm (GA) to mitigate the problems, but most of them are designed to optimize only one or two architectural factors of ANN. With this background, the paper presents a global optimization approach of ANN to predict the stock price index. In this study, GA optimizes multiple architectural factors and feature transformations of ANN to relieve the limitations of the conventional backpropagation algorithm synergistically. Experiments show our proposed model outperforms conventional approaches in the prediction of the stock price index.     

Air conditioner operation behaviour based on students' skin temperature in a classroom

A total of 25 college students participated in a study to determine when they would use an air conditioner during a lecture in a university classroom. The ambient temperature and relative humidity were measured 75 cm above the floor every minute. Skin temperatures were measured every minute at seven points, according to the recommendation of Hardy and Dubois. The average clothing insulation value (CLO) of subjects was 0.53 ± 0.07 CLO. The mean air velocity in the classroom was 0.13 ± 0.028 m/s. When the subjects turned the air conditioner both on and off, the average ambient temperatures, relative humidity and mean skin temperatures were 27.4 and 23.7 °C (p = 0.000), 40.9 and 40.0% (p = 0.528) and 32.7 and 32.2 °C (p = 0.024), respectively. When the status of the air conditioner was changed, the differences of skin temperatures in core body parts (head, abdomen and thigh) were not statistically significant. However, in the extremities (mid-lower arm, hand, shin and instep), the differences were statistically significant. Subjects preferred a fluctuating environment to a constant temperature condition. We found that a changing environment does not affect classroom study.     

Dynamic deformation behavior of ultrafine-grained low-carbon steels fabricated by equal-channel angular pressing

The dynamic deformation behavior of ultrafine-grained low-carbon steels fabricated by equal-channel angular pressing (ECAP) was investigated in this study. Dynamic torsional tests, using a torsional Kolsky bar, were conducted on four steel specimens, two of which were annealed at 480 °C after ECAP, and then the test data were compared in terms of microstructures, tensile properties, and adiabatic shear-band formation. The equal-channel angular pressed specimen consisted of very fine, equiaxed grains of 0.2 to 0.3 μm in size, which were slightly coarsened after annealing. The dynamic torsional test results indicated that maximum shear stress decreased with increasing annealing time, whereas fracture shear strain increased. Some adiabatic shear bands were observed at the gage center of the dynamically deformed torsional specimen. Their width was smaller in the equal-channel angular pressed specimen than in the 1-hour-annealed specimen, but they were not found in the 24-hour-annealed specimen. Ultrafine, equiaxed grains of 0.05 to 0.2 μm in size were formed inside the adiabatic shear band, and their boundaries had characteristics of high-angle grain boundaries. These phenomena were explained by dynamic recrystallization due to a highly localized plastic strain and temperature rise during dynamic deformation.