Balanced Split Research Articles

Deep learning based anime character sketch quality recognition

The explosion of the global anime market is poised to redefine the entertainment industry, boasting a valuation of USD 31.23 billion as of 2023 and projected to climb at a CAGR of 9.8% until 2030. At the heart of this growth is a vibrant community of artists and enthusiasts who navigate challenges such as artist scarcity and the lack of advanced tools for qualitative feedback and effective content promotion. Our study targets the core of anime creation sketching by evaluating the essential drawing elements and assessing their implementation in professional-quality anime portraits. We introduce an automated approach to predicting the anime quality using advances in deep learning. Utilizing transfer learning, we enhance three pre-trained models—MobileNetV2, ResNet50, and VGG16—with a customized dense layer, refining their capabilities for the binary classification of anime character sketches. The dataset employed comprises 155 images, categorized as 'Good' and 'Bad' to reflect the quality of sketches. A balanced split into training, validation, and testing subsets ensures a quantitative evaluation of data not seen previously by the model. The models, pre-trained on ImageNet and fine-tuned with our dataset, demonstrate varied sensitivity to hyperparameters, with the MobileNetV2 and ResNet50 model attaining a peak validation accuracy of 94% and the highest test accuracy of 79% indicating its potential as a robust tool for quality assessment in the anime industry. An interesting outcome of the research is that a lightweight MobileNetV2 model with much fewer parameters compared to other models resulted in the highest test accuracy.

SUSTAINABILITY IN PROPERTY TAXATION: AN APPLICATION FOR REVENUE STABILITY AND FAIRNESS PERSPECTIVES, THE CITY OF BUFFALO, NY

The scholarly modeling of property tax has always posed a challenge, with two primary concerns to be addressed: first, maintaining sustainability in the collection, which is primarily a concern for local governments; and second, ensuring fair distribution, which is of greater concern for citizens. In today’s practices, assessment-based property tax increases unmatched expenses in bubble economies. There is a substitution problem in rapid falls, the tendency to not decrease the assessments gives way to black holes and opens the door to ghost cities. This paper proposes alternative approaches, aside from market/land value or last sold price, aimed at improving sustainability and fairness rates. The dataset examined is based on 93.7K records and 88 attributes for assessed value of properties within the City of Buffalo, the United States of America. Since the label (Total Value) is a numerical and continuous value, regression models are selected, where ensemble machine learning methods categorically work well with larger datasets, combined with weak learners, like decision trees. Stacked Ensemble led the least error for regression with 0.98 R2, followed by Gradient Boosting. Results show a 79% dominance of uncontrollable attributes, such as Land Value, Neighborhood, and Sale (last sold) Price, compared to controllable attributes, such as Total Living Area, Construction Grade, Second Story Area, and many others. This article suggests having a more balanced split between uncontrollable and controllable attributes would contribute to both sustainability and fair distribution.

Large-Scale Modeling of Sparse Protein Kinase Activity Data.

Protein kinases are a protein family that plays an important role in several complex diseases such as cancer and cardiovascular and immunological diseases. Protein kinases have conserved ATP binding sites, which when targeted can lead to similar activities of inhibitors against different kinases. This can be exploited to create multitarget drugs. On the other hand, selectivity (lack of similar activities) is desirable in order to avoid toxicity issues. There is a vast amount of protein kinase activity data in the public domain, which can be used in many different ways. Multitask machine learning models are expected to excel for these kinds of data sets because they can learn from implicit correlations between tasks (in this case activities against a variety of kinases). However, multitask modeling of sparse data poses two major challenges: (i) creating a balanced train-test split without data leakage and (ii) handling missing data. In this work, we construct a protein kinase benchmark set composed of two balanced splits without data leakage, using random and dissimilarity-driven cluster-based mechanisms, respectively. This data set can be used for benchmarking and developing protein kinase activity prediction models. Overall, the performance on the dissimilarity-driven cluster-based split is lower than on random split-based sets for all models, indicating poor generalizability of models. Nevertheless, we show that multitask deep learning models, on this very sparse data set, outperform single-task deep learning and tree-based models. Finally, we demonstrate that data imputation does not improve the performance of (multitask) models on this benchmark set.

Physics-Informed Neural Networks with Group Contribution Methods.

Thermophysical properties of organic compounds are used in countless scientific, engineering, and industrial settings in developing theories, designing new systems and devices, analyzing costs and risks, and improving existing infrastructure. Often, due to costs, safety, prior interest, or procedural difficulties, experimental values for desired properties are not available and must be predicted. The literature is filled with prediction techniques, but even the best traditional methods have significant errors compared to what is possible considering experimental uncertainty. Recently, machine learning and artificial intelligence techniques have been applied to the property prediction problem, but the examples to date do not extrapolate well outside the data set used for training the model. This work demonstrates a solution to this problem by combining chemistry and physics when training the model and builds upon prior traditional and machine learning methods. Two case studies are presented. The first is for parachor which is used for surface tension prediction. Surface tensions are needed to design distillation columns, adsorption processes, gas-liquid reactors, liquid-liquid extractors, improve oil reservoir recovery, and undertake environmental impact studies or remediation actions. A set of 277 compounds is divided into training, validation, and test sets, and a multilayered physics-informed neural network (PINN) is developed. The results demonstrate that better extrapolation by deep learning models can be developed by adding in physics-based constraints. Second, a set of 1600 compounds is utilized for training, validating, and testing a PINN to improve normal boiling point predictions based on group contribution methods and physics-based constraints. The results show that the PINN performs better than any other method with a normal boiling point mean absolute error of 6.95 °C on training and 11.2 °C on test data. Key observations are that (1) a balanced split by compound type is important to have representative compound families in each of the train, validation, and test sets and (2) constraining group contributions being positive improves predictions on the test set. While this work demonstrates improvements for only surface tension and normal boiling point, the results offer significant hope that PINNs can improve prediction of other relevant thermophysical properties over existing approaches.

The Use of Code-Switching and Code-Mixing in English Learning for Children:

The research aimed to determine the types of communication applied in an intercultural family. They had diverse nationalities, customs, values, and beliefs. The researchers viewed communication as the main issue that could occur in an intercultural family. One of the examples of an intercultural family was Aprillia, a Makassarese woman married to Shawn Mullins, a U.S Military. They caused code-switching and code-mixing, types of code-switching and code-mixing applied in the family, children’s communication process, and strategies to solve communication problems in the intercultural family. The research’s finding was gathered from April’s family using an in-depth interview. By applying a qualitative approach, the researchers found the type of code-switching, code-mixing, and communication processes applied in April’s family. Furthermore, the researchers discover that techniques for resolving communication problems and linguistic barriers include avoiding ethnocentrism, challenging prejudices, recognizing differences, adjusting communication, equality, and a balanced split. Consequently, the most effective technique for resolving communication issues within an intercultural family is to remember that every individual has a distinct perspective. To avoid miscommunication during a conversation, individuals must learn to adapt to the capacity of the person with whom they are communicating.

On the study of teardrop shaped split and collision (TS-SAC) micromixers with balanced and unbalanced split of subchannels

ABSTRACT A teardrop-shaped split and collision (TS-SAC) micromixer based on the concept of balanced/unbalanced splits and collisions of fluid streams is presented. Numerical study is carried out for Reynolds numbers in the range of 0.1–75. In-depth flow and mixing dynamics for the micromixers are studied to assess mixing performance qualitatively. In order to produce unbalanced collision of the two fluid streams, difference in mass flow rate in the two sub-channels is utilized. The performance characteristics, viz., mixing index, pressure drop, mixing cost and mixing sensitivity are used to study mixing performance. A significant effect of unbalanced collisions of the fluid streams on the mixing enhancement is observed. Micromixer shows interesting mixing dynamics for different ratios sub-channels widths. A high mixing index and a low-pressure drop are observed for the micromixer with highest ratio of the widths of sub-channels. However, a low mixing index and a high-pressure drop are observed for the micromixer with lowest ratio of the widths of sub-channels.

Impact of megaships on the performance of port container terminals

Following the advent of megaships, the performance requirements of container terminals have increased significantly, highlighting necessary changes in their layout, infrastructure, and equipment. We focus on the impact of megaships on a terminal within the port network of the Italian Region of Liguria, in terms of its ability to manage the flow of imports from arrival to inland destinations. We use discrete event simulation techniques to analyze the operations of a terminal and evaluate the relevant performance indices in different scenarios, which vary as a function of the “call size” of the larger containerships. The possibility of guaranteeing a more balanced modal split (favoring rail transport) for the inland distribution of containers is also evaluated. Dwell times at the yard and turnaround times at the berth are considered, with the objective of achieving a modal split of inland transport consisting of no less than 40% rail. Our results show that this objective can be achieved if a higher dwell time for outgoing containers is allowed.

The place of VR technologies in UK architectural practice

ABSTRACTSince the year 2000, interactive immersive technology is widely affordable and available across disciplines and in the private domain. This paper presents the results of an original survey amongst the AJ100 (27 February 2018) practices in the UK exploring the use of VR technologies in contemporary architectural practice. It investigates the type and intensity of the use in practices different in size, type but unified by their high ambitions and recognition as the leading practices in the UK. The survey reveals that 52% of the AJ100 practices use VR in their offices. Sixty percent of the VR outputs are 360-degree panoramic images leaving 40% to (Construct UK Ltd) the production of VR walkthrough models. The surveyed practices report an almost balanced split between desktop tethered hardware and mobile devices, whereas large hardware installations i.e. CAVE VR or Space Domes play a subordinate role in their office equipment. It is not surprising that the VR technologies are heavily used in client meetings and for representation purposes, but the study also revealed that the VR has found its way into the genuine design processes within the core design team as well as in coordination activities between stakeholders outside the client related communication, i.e. consultant communication. The findings of the study are based on a response rate of 72% including 13 offices from the top 20 of the AJ100 practices in the UK.

Flow field analysis of a passive wavy micromixer with CSAR and ESAR elements

The performance analysis of wavy micromixers with split and recombine (SAR) elements has been carried out. The two types of SAR elements viz. circular split and recombine elements and elliptical split and recombine elements are used to get benefit of SAR mechanism. The main concept in the proposed design is to enhance the interfacial area between the two fluids by creating a transverse flow and split and recombination of fluid flow streams with the help of SAR elements. The effects of balanced split and unbalanced split of SAR elements and operational parameter Reynolds number on the mixing index and pressure drop are evaluated. The COMSOL Multiphysics 5.0 is used for computational fluid dynamics analysis of micromixers. The analysis is carried out for balanced and unbalanced split of the mixing elements at various Reynolds numbers in the range of 0.1–75. The computation interfacial patterns on the planes perpendicular to the direction of fluid flow, mixing index and pressure drop are studied in depth. The results indicate that at inlet Reynolds number below 5, the mixing performance is diffusion dominated. However at Reynolds number greater than 5, the mixing index increases intensely due secondary flow and separation vortices, as well as SAR effect.

Optimum earthquake-tuned TMDs: Seismic performance and new design concept of balance of split effective modal masses

The present paper, concerning the performance of optimum TMDs in seismic engineering, is composed of two parts. First, Part I deals with a comprehensive investigation on the effectiveness of an optimum seismic-tuned TMD towards reducing the earthquake response of linear frame structures, which is set up by a dedicated optimisation algorithm, allowing to achieve the best control device for a specific structural system and earthquake event. First, a representative set of 16 primary structures and 18 selected earthquakes has been considered, for a total of 288 cases. Second, additional trials based on a benchmark 10-storey frame and on a codified set of far-field earthquakes (FEMA P695 database), composed of 2×22=44 records, have been developed, for a further assessment. The presented outcomes provide extensive information about the TMD performance in the seismic engineering scenario, for a-priori known seismic input. Then, Part II reports a comprehensive post-processing and interpretation analysis of the results above. Specifically, connections between seismic TMD performance and modal structural properties are envisaged. Cross-comparisons, presented here through typical sample cases, analyse the changes of the structural characteristics after TMD insertion and monitor the associated effects on the seismic response reduction. The crucial role of achieving a balanced split of effective modal masses on the tuned mode is investigated, and highlighted as an essential ingredient to rely on a high TMD effectiveness in the seismic context. This criterion may be implemented within the tuning process, as a basic true structure-based concept for TMD design in Earthquake Engineering applications, thus for a-priori unknown seismic input.

Experimental measurements of gas–solid flow and splitting mechanisms of a coal pipe splitter with a perpendicularly arranged upstream elbow

The effect of the vertical pipe length on the performance of a coal pipe splitter with a perpendicularly arranged upstream elbow was investigated experimentally employing a fiber optic measuring system. The upstream elbow and coal pipe splitter were installed in two perpendicular planes. Contours of distributions of the particle concentration and size were obtained in different transverse sections. The experimental data show that the maximum/minimum concentration ratio in transverse sections A, B, and C decreased rapidly as the length of the vertical pipe increased. The left/right-leg average concentration ratio remained about 1, and a balanced split was thus achieved. With a perpendicularly arranged upstream elbow, the vertical pipe length had little effect on the splitter performance, which is beneficial for engineering design and convenient for industrial application.

3D Ear Segmentation and Classification Through Indexing

Current growth trends in different biometrics applications present challenges to researchers. To address these challenges, we need new data storage and retrieval techniques to make the recognition process time efficient. This paper presents a system for time efficient 3D ear biometrics from a large biometrics database. The proposed system has two components that are primarily responsible for: 1) automatic 3D ear segmentation and 2) hierarchical categorization of the 3D ear database using the shape information and surface depth information, respectively. We use an active contour algorithm along with a tree-structured graph to segment the ear region from the 3D profile images. The segmented 3D ear database is then categorized based on the geometrical feature values, computed from the ear shape, into oval, round, rectangular, and triangular categories. For the categorization based on the depth information, the feature space is partitioned using tree-based indexing techniques. We used indexing techniques with balanced split (k-dimensional (KD) tree) and unbalanced split (pyramid tree) data structures to categorize the database separately and then compared their retrieval efficiency. Experiments are conducted to compare the average computation time per query when performing recognition through hierarchical categorization with the average computation time when recognition is based on sequential search. Experimental results without indexing conducted on the University of Notre Dame Collection J2 data set yielded a rank-one recognition rate of 98.5%. We applied the indexing technique to compute the rank-one recognition accuracy with a 10%–50% search space reduction using a 10% step size. With 10%, 20%, 30%, 40%, and 50% search space reduction the rank-one recognition accuracy gracefully degrades to 96.87%, 96.14%, 95.18%, 94.21%, and 93.49%, respectively, while performing nearly 3, 3.3, 4, 4.2, and 5 times faster than the state-of-the-art technique that uses sequential search.

Evaluation of Infrastructure Investments: The Case of the Southern Italian Transport Network

The EU Community Support Framework (CSF), in the transport sector, aims at implementing a sustainable network characterised by a more balanced modal split and by integrated connectivity among the transport nodes. This study provides a concise description of the CSF, PON and POR programmes and tests - by means of a multicriteria approach - the degree of coherence among them. The results highlight that the strategy pursued during the 2001-2006 period sought to strengthen some priority nodes located in the TEN-T network and that there is a marked heterogeneity between the PON projects and the POR projects.

Balanced split sets and Hamilton–Jacobi equations

We study the singular set of solutions to Hamilton–Jacobi equations with a Hamiltonian independent of u. In a previous paper, we proved that the singular set is what we called a balanced split locus. In this paper, we find and classify all balanced split loci, identifying the cases where the only balanced split locus is the singular locus, and the cases where this does not hold. This clarifies the relationship between viscosity solutions and the classical approach of characteristics, providing equations for the singular set. Along the way, we prove more structure results about the singular sets.

Efficient white organic light‐emitting diodes based on a balanced split of the exciton‐recombination zone using a graded mixed layer as an electron‐blocking layer

Abstract— Efficient white organic light‐emitting diodes with both a graded mixed layer as the blue‐emitting layer and an electron‐blocking layer, and a DPVBi:Rubrene layer as a yellow‐emitting layer have been demonstrated. The mixing of the two colors occurs due to a balanced split of the exciton‐recombination zone by the graded mixed layer serving as the electron‐blocking layer. The white organic light‐emitting diode with an ITO/2‐TNATA 30 nm/NPB 30 nm/DPVBi:Rubrene (1.0 wt.%) 5 nm/NPB:DPVBi (9:1) 150 nm/NPB:DPVBi (5:5) 75 nm/NPB:DPVBi (3:7) 75 nm/NPB:DPVBi (2:8) 75 nm/NPB:DPVBi (0.5:9.5) 75 nm/BCP 5 nm/Alq3 30 nm/LiF 0.5 nm/Al 100 nm structure is chosen as a device with an optimal configuration among devices investigated in this study. The employment of the graded mixed layer in the device is effective in suppressing the color shift at different voltages. The white light, with a Commission Internationale d'Eclairage chromaticity coordinates of (0.33, 0.34), is obtained with an applied voltage of 10.5 V for the device. At the applied voltage, the luminance is 4882 cd/m2 and the current efficiency is 5.03 cd/A.

Changes in the Dynamic Relationship between ER and HER2 in Paired Breast Cancer (BC) Biopsies from Patients (Pts) with Recurrent Tamoxifen-Resistant Breast Cancer.

Abstract Background: Paired biopsies at diagnosis and relapse on tamoxifen (TAM) provide information regarding adaptive changes associated with endocrine resistance. We previously reported that conversion to a HER2+ phenotype occurs infrequently1, however modest increases in HER2 may be relevant in the context of TAM resistance. Also while TAM resistance in general is associated with a fall in mean ER and PgR levels, this masks changes in individual pts.Aim: To determine whether a subset of HER2 negative BC demonstrated modest increases in HER2 by IHC and FISH, and if so whether this correlated with changes in ER and PgR expression in individual pts.Methods: Triplicate TMAs were constructed of paired primary and recurrent tumor samples from pts relapsing after adjuvant TAM. TMAs were stained for ER and HER2, whole sections for PgR. HER2 amplification was assessed by FISH.Results: Paired samples from primary and TAM-resistant recurrences were available on 53 ER+ and/or PgR+ pts. 10% had a HER2+ primary (IHC3+ and FISH+). 19% (10/53) showed increasing HER2 in the relapsed tumor (0 to 1+ in 5 pts, 1+ to 2+ in 2 pts and 0/1+ to 3+ in 3 pts). The 3 pts converting to 3+ also became FISH+. 71% (38/53) had stable or decreasing HER2. DFI was 50.3 months in the subset with stable or decreasing HER2, 40.2 in the subset with increasing HER2, and 35.2 in the HER2+ subset (p=NS). We next compared changes in ER and PgR in the 3 subsets. Among pts with stable or decreasing HER2 (N=38), there was a significant decrease in mean H-scores for both ER (p=0.001) and PgR (p=0.002) at relapse (Fig 1); the majority recurred with reduced ER (76%) or PgR levels (84%). In contrast, among pts with increasing HER2, there was no change in mean ER or PgR levels (Fig 2A); however when considering changes in individual pts, there was a balanced split between tumors relapsing with up- vs. down-regulation in ER/PgR (Fig 2B,C). In the HER2 positive subset, 100% of the pts relapsed with suppressed ER and PgR.Conclusion: In an updated series of TAM-resistant pts, we show that while conversions to HER2 3+/FISH+ remain rare (5%), modest increases in HER2 are more frequent (19%). TAM resistance in general is associated with decreasing ER and PgR. However, among pts with adaptive increases in HER2, TAM resistance is associated with either down- or up-regulation in ER expression and function. This is consistent with data from acquired endocrine resistance models we previously reported2,3.1Gutierrez et al, JCO 20052Pancholi et al, ERC 2008,3 Martin et al, JBC 2003 Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 704.

Environmentally Sustainable Transport in the CEI Countries

Moving people and freight in an environmentally sustainable manner that reduces environmental pollution and health hazards is a key challenge for transport and environment policies in Europe. Present mobility patterns of passenger and freight transport in Central and Eastern Europe do not correspond to the objectives of sustainable development. This paper presents the results of a transport futures study for the CEI region as a whole using backcasting methodology with long-term sustainability criteria to be met by 2030. Achieving environmentally sustainable transport (EST) doesn’t mean less transport and mobility than we have today, but it means primarily maintaining a balanced modal split that results in less environmental and health impacts than it would be under projected future trends. Rail, trams, busses and new forms of flexible inter-modal public transport mobility would have to take a large share and rail transport for passenger and freight as well as inland shipping would have to be nearly doubled by 2030 while road freight could still increase if it is based on alternative fuels reducing its impacts. Technological advancements for passenger cars and lorries, fuels and infrastructure will play an important part to achieve EST, but also ‘smart’ mobility management (e.g. transport avoidance, increasing load factors and modal shift), innovative mobility services and freight logistics would be critical. The implementation of these policies and strategies will require coherent and comprehensive packages of instruments and measures, including: economic instruments, regulatory instruments, changes in infrastructure investment, mobility management, information and education programmes as well as better integration of land use, transport and environment policies. Realising EST will provide new opportunities for businesses to develop and invest in innovative solutions for passenger and freight transport. Overall, achieving EST would constitute a net benefit for the environment and for quality of life in general.

Bone healing capacity of titanium plasma-sprayed and hydroxylapatite-coated oral implants.

The influence of surface quality, in particular surface topography and implant material, was evaluated by inserting titanium- and hydroxylapatite plasma-sprayed coated implants into the maxilla of 10 goats. Three types of plasma-spray coatings were applied to tapered, screw shaped implants; titanium plasma-spray coating (TPS), titanium plasma-spray coating with additional acid passivation (TPSA) and a bilayered coating (TPS/HA) consisting of titanium plasma-spray coating (TPS) and a hydroxylapatite part (HA). In addition, as-machined implants (TiM) were used as control. A total of 40 implants were inserted according a balanced split plot design. At the end of a 3-month healing period, it appeared that 5 implants (2 TPS, 1 TPSA, 1 TPS/HA and 1 TiM) were lost. Histological examination revealed a stronger bone response to TPS/HA coated implants. Even the TPS/HA coated implants induced bone formation on the part of the implant inserted into the sinus. No signs of delamination of the TPS coatings were visible. The HA part of the dual coating showed signs of degradation. Histomorphometrical analysis confirmed these findings. A significant difference in bone contact (P < 0.05) was measured between the TPS/HA coated implants and the other types of implants. Linear regression (r = 0.27) showed no correlation between the inscrew values at the base line and the bone contact measurements 3 months after healing. On the basis of these results, we can conclude that the chemical composition of the HA coating has a positive influence on the bone reaction. The influence of roughness is less evident.