Bidirectional Training Research Articles

Community Leadership Institute for Equity: Planning Processes and Procedures to Develop Partnered Conferences

Abstract: Background: Community-partnered participatory research (CPPR) is a research approach that supports equitable collaboration of community and academic co-leaders in research and policy. Despite CPPR’s 25-year history, infrastructure supporting community members in bidirectional learning has not been formalized. Objective: This paper describes processes and procedures using CPPR to plan conferences to develop community leadership training infrastructure. Methods: We utilized rapid ethnographic analysis to examine conference planning processes for community leadership in CPPR. Community and academic leaders in Los Angeles, New Orleans, and Chicago met weekly over two months to plan, given COVID-19, three Zoom conferences on a leadership training institute for CPPR, with planning for (1) community co-leadership in research and policy; (2) local and national CPPR programs; and (3) models for bidirectional training. Results: The planning process emphasized bidirectional learning for community and academic members for research and services/policy to benefit communities, within a Community Leadership Institute for Equity (C-LIFE) to promote equity and power sharing for community leaders. The planning process identified major themes of framing of C-LIFE conference planning goals, developing the conference structure, promoting equity and diversity, envisioning the future of CPPR, challenges, collaborations, future curriculum ideas for C-LIFE, evaluation and next-steps for Zoom conferences in November 2020. Conclusions: It was feasible to use CPPR to plan Zoom conferences to promote community leadership training across multiple sites. Key planning themes included promoting equity, addressing structural racism, bidirectional learning and integrating community, academic, and policy priorities with community co-leaders as change agents.

Determining community happiness index with transformers and attention-based deep learning

In the current digital era, evaluating the quality of people's lives and their happiness index is closely related to their expressions and opinions on Twitter social media. Measuring population welfare goes beyond monetary aspects, focusing more on subjective well-being, and sentiment analysis helps evaluate people's perceptions of happiness aspects. Aspect-based sentiment analysis (ABSA) effectively identifies sentiments on predetermined aspects. The previous study has used Word-to-Vector (Word2Vec) and long short-term memory (LSTM) methods with or without attention mechanism (AM) to solve ABSA cases. However, the problem with the previous study is that Word2Vec has the disadvantage of being unable to handle the context of words in a sentence. Therefore, this study will address the problem with bidirectional encoder representations from transformers (BERT), which has the advantage of performing bidirectional training. Bayesian optimization as a hyperparameter tuning technique is used to find the best combination of parameters during the training process. Here we show that BERT-LSTM-AM outperforms the Word2Vec-LSTM-AM model in predicting aspect and sentiment. Furthermore, we found that BERT is the best state-of-the-art embedding technique for representing words in a sentence. Our results demonstrate how BERT as an embedding technique can significantly improve the model performance over Word2Vec.

International collaborative research, systems leadership and education: reflections from academic biomedical researchers in Africa

ScopeAcademic biomedical researchers and educators in African countries navigate complex local, national, and international systems to conduct grant-funded research. To secure funding, collaboration with researchers from high-income countries is often necessary. Existing literature highlights that these global health initiatives are commonly fraught with unequal power dynamics and lead by the foreign partners. Despite these inequalities, African faculty can benefit from these collaborations, fostering the development of their labs and careers. This study delves into reflections on lived experiences from academic biomedical researchers in Africa to better understand the impact of foreign collaborations.MethodsWe designed a qualitative study using the Interpretative Phenomenological Analysis (IPA) method and used Self-determination and Complex Systems Leadership theories to frame this study. Ten academic biomedical researchers in Africa consented to join this study. The participants submitted a four-week series of reflective journals through an online data management platform. Subsequently, IPA methods were employed to analyze the collected journals.ResultsParticipants’ reflections yielded six thematic key findings, encompassing their experiences in international collaborative research. The findings included: foreign dominance within the international macrosystem; resource challenges in their local microsystems; mesosystem dependency from collaborations; scholarly identity within research relationships; collaborative leadership; and the impact of the foreign perspective. From these findings, three implications were drawn suggesting that participants were (1) motivated by scholarly identity found in collaborations; (2) depended on collaborations that were colonialist but beneficial; and (3) created change through leadership at the microsystem level.ImplicationsForeign collaborators and funders in global health education and research should critically consider how implications of this study relate to their collaborative work. Based on our analysis, recommendations for foreign collaborators and funders include prioritizing local leadership and perspectives in education initiatives and research grants; reviewing and leveraging collective leadership; engaging in bidirectional training, and mentoring opportunities; participating in power assessments; and removing publication barriers for researchers in Africa (and low-and middle-income countries). Insights from this study could impact global health research and education in multiple ways including new initiatives in mentorship and training in international collaborations along with increased awareness and correction of colonialism within these collaborations.

Neural session key exchange in the Industrial Internet of Things using hyperchaotic-guided vector-valued artificial neural synchronization

The Industrial Internet of Things’ (IIoT) ongoing growth has provided engineers with a wealth of prospects for improving machine throughput. Due to advancement, owing to the unreliable aspect of the communication route, many business managers still hesitate to operate their equipment online. Security is critical in the IIoT, as certain IIoT devices gather sensory data for vital societal production and living. If the entities are authenticated and confidence is established, the use of the Internet to manage industrial activities can become widely embraced. The hyperchaotic neuronal coordinated vector-valued key exchange presented in this study enables safe Wireless Sensor Network (WSN) transmission in the IIoT. Key coordination in IIoT is the main issue. As a part of the security protocols for IIoT, this study presented a vector-valued neuronal coordination-based key synchronization. How well the ANNs on the collaborating edges synchronize instead of each other’s weights is a key factor in neural coordination. Existing techniques take too long to evaluate collaboration, endangering the secrecy of neural synchronization. In addition, minimal studies have been conducted on the reciprocal training of a set of ANNs and the use of a robust PRNG to provide a common input. The suggested approach has a lot of benefits: (1) This research describes the fast evaluation of the coordination of a group of ANNs using a hyperchaotic framework to coordinate ANNs for key agreement. (2) ANNs are coordinated via bidirectional training to spread the neuronal key over a communication link. (3) The measure in which the ANNs have provided similar outputs in previous rounds is used to determine synchronization. (4) The input sequencing for ANN are produced using a collaboratively created 4D, 5D, and 6D hyperchaotic-guided PRNG. (5) The IIoT network’s session key is created via a reciprocating neuronal synchronization of vector-valued ANNs. (6) In compared to earlier methodologies, the proposed technique accelerates the procedure through which participated ANNs achieve perfect synchronization. The suggested approach works better than comparable methods that have been published.

Florida-California Cancer Research, Education and Engagement (CaRE2) Health Equity Center: Structure, Innovations, and Initial Outcomes.

The Florida-California Cancer Research, Education, and Engagement (CaRE2) Health Equity Center is a triad partnership committed to increasing institutional capacity for cancer disparity research, the diversity of the cancer workforce, and community empowerment. This article provides an overview of the structure, process innovations, and initial outcomes from the first 4years of the CaRE2 triad partnership. CaRE2 serves diverse populations in Florida and California using a "molecule to the community and back" model. We prioritize research on the complex intersection of biological, environmental, and social determinants health, working together with scientific and health disparities communities, sharing expertise across institutions, bidirectional training, and community outreach. Partnership progress and outcomes were assessed using mixed methods and four Program Steering Committee meetings. Research capacity was increased through development of a Living Repository of 81 cancer model systems from minority patients for novel cancer drug development. CaRE2 funded 15 scientific projects resulting in 38 publications. Workforce diversity entailed supporting 94 cancer trainees (92 URM) and 34 ESIs (32 URM) who coauthored 313 CaRE2-related publications and received 48 grants. Community empowerment was promoted via outreaching to more than 3000 individuals, training 145 community cancer advocates (including 28 Community Scientist Advocates), and publishing 10 community reports. CaRE2 members and trainees together have published 639 articles, received 61 grants, and 57 awards. The CaRE2 partnership has achieved its initial aims. Infrastructure for translational cancer research was expanded at one partner institution, and cancer disparities research was expanded at the two cancer centers.

Claim Analyzer : Evaluating Credibility of Data

The current state of news in the era of computers The news environment in the modern computer age The news environment in the modern computer age The news environment in the modern computer age. The news environment in the modern computer age Social media platforms have taken the role of the antiquated traditional print media as a part of the news ecosystem in the current computer era. False news is spread at an astounding velocity and scale because social media platforms allow us to consume news much more rapidly and with less restricted editing. According to recent research, several efficient methods for spotting fake news encrypt social context-level data and news content sequential neural networks using a unidirectional examination of the text sequence. In order to represent the pertinent information of false news and improve classification performance while capturing semantic and long-distance connections in sentences, a bidirectional training strategy is a necessity. Claim Analyzer is only a model for evaluating the veracity of claims made online The statements might either be True or False. Fake news is a kind of propaganda in which false information is knowingly disseminated via news organizations and/or social media platforms. It is crucial to create methods of spotting false news information since its spread can have detrimental effects, such as influencing elections and widening political rifts. BERT is intended to simultaneously condition on both left and right context in all layers in order to pre-train deep bidirectional representations from unlabeled text. Therefore, state-of-the-art models for a variety of tasks, including question answering and language inference, may be created using just one extra output layer to fine-tune the pre-trained BERT model without making significant task-specific architecture adjustments. By merging several parallel blocks of the single-layer deep Convolutional Neural Network (CNN)with the BERT and variable kernel sizes and filters, we propose a BERT-based (Bidirectional Encoder Representations from Transformers) deep learning technique. The biggest obstacle to natural language understanding is ambiguity, which may be handled with the help of this combination. When applied to huge datasets, our suggested model provides88% accuracy.

RMCL: A deep learning based recursive malicious context learner in social networks

AbstractPropagation of malicious content and its promotion is a recurrent problem prominent in social networks. Detection and interactive labeling of malicious context promotion on‐the‐fly is a very challenging and difficult task due to the lack of complete knowledge about the underlying context. Modern research in this field use neural networks and text encoders to analyze the information dynamically which is inefficient, in terms of time‐space consumed in the overall process. This article proposes an online active learning system to label the data streams on‐the‐fly by sampling them in the form of tweet‐retweet‐follow (trf) sequences in social networks. A heuristically pretrained recursive malicious context learner is fixed for knowledge acquisition, data accommodation and pseudonymization in the form of a transformer in the tree structured recursive neural network. The data stream is trained using recursive bidirectional training to capture long‐term dependencies. The COVID‐19 data streams are selectively sampled on‐the‐fly using psycho linguistic words in the proposed experiment and are labeled deceptive/nondeceptive based on knowledge learned during pretraining. The proposed recursive malicious context learner successfully resolves the problem of on‐the‐fly‐interactive labeling of a dynamically changing data stream.

Enhancing Integrity Modeling for Emotional Conversation Generation

Establishing an open-domain emotional conversation generation with content and emotional dependence is a key step toward intelligent interactions. However, the researchers only dedicated to adding the thematic content or emotional words and handling both factors are not yet properly solved. In this article, we aim to address the issue of integrity in emotional content which indicates the response has a dependency relationship between theme and emotion. We propose a thematic–emotional conversation model, which consists of parallel channels and joint encoding, to maximize integrity in conversation generation. We first use the encoding of the parallel channel to strengthen the decoding weight of the theme and emotion. The joint encoding of the dual attention mechanism and the thematic emotional tradeoff strategy are involved in the encoding to ensure that the theme complements the emotional content during the response generation. After that, the bidirectional training interactions between post and response are calculated to derive the new response inference vectors. Experiment results suggest that the proposed model is effective not only in content integrity but also in emotion.

Assessment of Running Safety of Developed Train–Bridge–Wave Model Under One-Direction and Two Bidirectional High-Speed Train Passing Subjected to Rail Irregularities

A study about the running safety of trains moving over bridges subjected to hydrodynamic loads under one-direction and two bidirectional high-speed train passing is presented. Dynamic performance of the developed 3D train–bridge–wave system (TBW model) subjected to different hydrodynamic loads and applying Association of American Railroads track irregularities is established in this study. The dynamic responses of the bridge, which is under one-direction and two bidirectional high-speed train passing and subjected to several hydrodynamic loads and rail irregularities, are analyzed. The dynamic responses of the bridge and the running safety indices of the trains on the bridge under several types of sea states in one-direction and two bidirectional train passing are analyzed. Results of developed TBW’s sensitivity analyses have shown the importance of sea-state conditions for safe train running. Irregularities have an obvious effect on the dynamic responses of the bridge. Between comparing one-direction and two bidirectional high-speed train passing modes, bidirectional train passing has a greater effect on vertical acceleration and displacement than horizontal ones in the presence of a hydrodynamic load. Therefore, the combination of wave forces, rail irregularities, and bidirectional train passing jeopardizes the running safety of high-speed trains by increasing both lateral and vertical displacements and accelerations. Furthermore, the effect of wave force on the running safety indices of the train is investigated by considering different train speeds. As expected, the tendency observed in all the derailment criteria is similar, indicating that the risk of derailment increases with the increasing of the running speed and wave height.

Parallel Bidirectionally Pretrained Taggers as Feature Generators

In a setting where multiple automatic annotation approaches coexist and advance separately but none completely solve a specific problem, the key might be in their combination and integration. This paper outlines a scalable architecture for Part-of-Speech tagging using multiple standalone annotation systems as feature generators for a stacked classifier. It also explores automatic resource expansion via dataset augmentation and bidirectional training in order to increase the number of taggers and to maximize the impact of the composite system, which is especially viable for low-resource languages. We demonstrate the approach on a preannotated dataset for Serbian using nested cross-validation to test and compare standalone and composite taggers. Based on the results, we conclude that given a limited training dataset, there is a payoff from cutting a percentage of the initial training set and using it to fine-tune a machine-learning-based stacked classifier, especially if it is trained bidirectionally. Moreover, we found a measurable impact on the usage of multiple tagsets to scale-up the architecture further through transfer learning methods.

Development of Clinical Practice in Adult Nursing (Chronic Care) During the COVID-19 Pandemic - Building and Implementing an Online Training Program

The Department of Nursing, School of Health Sciences, University of Occupational and Environmental Health provides the specialized course "Clinical Practice in Adult Nursing (Chronic Care)" as a compulsory subject for third- and fourth-year students to acquire the basic skills required in nursing practice for adult patients with chronic diseases. Students are trained in hospital wards and in outpatient medical departments/admission support units for a total of 3 weeks to communicate directly with patients with chronic diseases, and to deepen their learning about the specialty of chronic nursing care. The COVID-19 pandemic forced the Department of Nursing to hold prompt discussions and determine an appropriate method to provide clinical training under this situation. It decided to cancel the conventional face-to-face training program, which was scheduled to start in May 2020, and to implement an online training program instead. Remote training was provided using the online conferencing system Zoom. The online training program was designed to realize simultaneous bidirectional training. To prepare for the training, simulated cases and simulated medical records were created. Faculty members played the roles of "simulated patients", "clinical supervisors", and "faculty members", while adopting elaborate approaches to help students develop nursing practice in real time. As a result, students were able to participate in training with a real feeling, even in an online environment, as they communicated with simulated patients daily, similarly to the case of clinical training. They also realized the limitations of learning without using the five physial senses and providing patient support with appropriate timing, revealing challenges specific to online training.

Learning Dense and Continuous Optical Flow From an Event Camera.

Event cameras such as DAVIS can simultaneously output high temporal resolution events and low frame-rate intensity images, which own great potential in capturing scene motion, such as optical flow estimation. Most of the existing optical flow estimation methods are based on two consecutive image frames and can only estimate discrete flow at a fixed time interval. Previous work has shown that continuous flow estimation can be achieved by changing the quantities or time intervals of events. However, they are difficult to estimate reliable dense flow, especially in the regions without any triggered events. In this paper, we propose a novel deep learning-based dense and continuous optical flow estimation framework from a single image with event streams, which facilitates the accurate perception of high-speed motion. Specifically, we first propose an event-image fusion and correlation module to effectively exploit the internal motion from two different modalities of data. Then we propose an iterative update network structure with bidirectional training for optical flow prediction. Therefore, our model can estimate reliable dense flow as two-frame-based methods, as well as estimate temporal continuous flow as event-based methods. Extensive experimental results on both synthetic and real captured datasets demonstrate that our model outperforms existing event-based state-of-the-art methods and our designed baselines for accurate dense and continuous optical flow estimation.

DeepCollaboration: Collaborative Generative and Discriminative Models for Class Incremental Learning

An important challenge for neural networks is to learn incrementally, i.e., learn new classes without catastrophic forgetting. To overcome this problem, generative replay technique has been suggested, which can generate samples belonging to learned classes while learning new ones. However, such generative models usually suffer from increased distribution mismatch between the generated and original samples along the learning process. In this work, we propose DeepCollaboration (D-Collab), a collaborative framework of deep generative and discriminative models to solve this problem effectively. We develop a discriminative learning model to incrementally update the latent feature space for continual classification. At the same time, a generative model is introduced to achieve conditional generation using the latent feature distribution produced by the discriminative model. Importantly, the generative and discriminative models are connected through bidirectional training to enforce cycle-consistency of mappings between feature and image domains. Furthermore, a domain alignment module is used to eliminate the divergence between the feature distributions of generated images and real ones. This module together with the discriminative model can perform effective sample mining to facilitate incremental learning. Extensive experiments on several visual recognition datasets show that our system can achieve state-of-the-art performance.

Bi-Directional Training Methods With Frequency-Division Duplexing

We study distributed algorithms for joint adaptation of precoding and combining filters in frequency division duplex (FDD) multiple-input multiple-output (MIMO) cellular systems. Our approach extends bi-directional training (BiT), designed for time division duplex (TDD) systems, to FDD systems where uplink/downlink reciprocity may not apply. An analysis of the performance loss due to different uplink-downlink frequencies in a point-to-point scenario, shows that a direct application of BiT gives mismatched precoders with substantial performance degradation. We first propose an algorithm assuming each base transceiver station (BTS) knows the uplink and downlink channels of mobiles within the cell. We then consider the scenario where neither the BTSs nor the mobiles have a priori channel state information. Our proposed approaches assume angular reciprocity characterized by angles of arrival/departure that vary predictably with frequency. Hence spatial beams corresponding to angles of arrival can be turned around to the corresponding angles of departure in the paired band. We present three methods, differing in how the angular decomposition is applied and used, namely, to reconstruct the combiner directly, or to reconstruct the received signal. Simulation results indicate that when the multipath is sufficiently sparse, most of achievable gain with channel reciprocity and TDD can be recovered.

A deep learning approach with temporal consistency for automatic myocardial segmentation of quantitative myocardial contrast echocardiography.

Quantitative myocardial contrast echocardiography (MCE) has been proved to be valuable in detecting myocardial ischemia. During quantitative MCE analysis, myocardial segmentation is a critical step in determining accurate region of interests (ROIs). However, traditional myocardial segmentation mainly relies on manual tracing of myocardial contours, which is time-consuming and laborious. To solve this problem, we propose a fully automatic myocardial segmentation framework that can segment myocardial regions in MCE accurately without human intervention. A total of 100 patients' MCE sequences were divided into a training set and a test set according to a 7: 3 proportion for analysis. We proposed a bi-directional training schema, which incorporated temporal information of forward and backward direction among frames in MCE sequences to ensure temporal consistency by combining convolutional neural network with recurrent neural network. Experiment results demonstrated that compared with a traditional segmentation model (U-net) and the model considering only forward temporal information (U-net + forward), our framework achieved the highest segmentation precision in Dice coefficient (U-net vs U-net + forward vs our framework: 0.78 ± 0.07 vs 0.79 ± 0.07 vs 0.81 ± 0.07, p < 0.01), Intersection over Union (0.65 ± 0.09 vs 0.66 ± 0.09 vs 0.68 ± 0.09, p < 0.01), and lowest Hausdorff Distance (32.68 ± 14.6 vs 28.69 ± 13.18 vs 27.59 ± 12.82 pixel point, p < 0.01). In the visual grading study, the performance of our framework was the best among these three models (52.47 ± 4.29 vs 54.53 ± 5.10 vs 57.30 ± 4.73, p < 0.01). A case report on a randomly selected subject for perfusion analysis showed that the perfusion parameters generated by using myocardial segmentation of our proposed framework were similar to that of the expert annotation. The proposed framework could generate more precise myocardial segmentation when compared with traditional methods. The perfusion parameters generated by these myocardial segmentations have a good similarity to that of manual annotation, suggesting that it has the potential to be utilized in routine clinical practice.

FakeBERT: Fake news detection in social media with a BERT-based deep learning approach

In the modern era of computing, the news ecosystem has transformed from old traditional print media to social media outlets. Social media platforms allow us to consume news much faster, with less restricted editing results in the spread of fake news at an incredible pace and scale. In recent researches, many useful methods for fake news detection employ sequential neural networks to encode news content and social context-level information where the text sequence was analyzed in a unidirectional way. Therefore, a bidirectional training approach is a priority for modelling the relevant information of fake news that is capable of improving the classification performance with the ability to capture semantic and long-distance dependencies in sentences. In this paper, we propose a BERT-based (Bidirectional Encoder Representations from Transformers) deep learning approach (FakeBERT) by combining different parallel blocks of the single-layer deep Convolutional Neural Network (CNN) having different kernel sizes and filters with the BERT. Such a combination is useful to handle ambiguity, which is the greatest challenge to natural language understanding. Classification results demonstrate that our proposed model (FakeBERT) outperforms the existing models with an accuracy of 98.90%.

Bidirectional Associative Memories: Unsupervised Hebbian Learning to Bidirectional Backpropagation

Bidirectional associative memories (BAMs) pass neural signals forward and backward through the same web of synapses. Earlier BAMs had no hidden neurons and did not use supervised learning. They tuned their synaptic weights with unsupervised Hebbian or competitive learning. Two-layer feedback BAMs always converge to fixed-point equilibria for threshold or threshold-like neurons. Every rectangular connection matrix is bidirectionally stable. These simpler BAMs extend to arbitrary hidden layers with supervised learning if the resulting bidirectional backpropagation algorithm uses the proper layer likelihood in the forward and backward directions. Bidirectional backpropagation lets users run deep classifiers and regressors in reverse as well as forward. Bidirectional training exploits pattern and synaptic information that forward-only running ignores.

BFRVSR: A Bidirectional Frame Recurrent Method for Video Super-Resolution

Video super-resolution is a challenging task. One possible solution, called the sliding window method, tries to divide the generation of high-resolution video sequences into independent subtasks. Another popular method, named the recurrent algorithm, utilizes the generated high-resolution images of previous frames to generate the high-resolution image. However, both methods have some unavoidable disadvantages. The former method usually leads to bad temporal consistency and has higher computational cost, while the latter method cannot always make full use of information contained by optical flow or any other calculated features. Thus, more investigations need to be done to explore the balance between these two methods. In this work, a bidirectional frame recurrent video super-resolution method is proposed. To be specific, reverse training is proposed that also utilizes a generated high-resolution frame to help estimate the high-resolution version of the former frame. The bidirectional recurrent method guarantees temporal consistency and also makes full use of the adjacent information due to the bidirectional training operation, while the computational cost is acceptable. Experimental results demonstrate that the bidirectional super-resolution framework gives remarkable performance and it solves time-related problems.

An ethics-based approach to global health research part 1: Building partnerships in global health

Global health partnerships (GHPs) can be the cornerstone for advancing research and public health. The steps to build a global research partnership focus on sharing a common research agenda, identifying key partners in the community, and establishing goals and expectations for partnerships. Moreover, upholding important values, such as communication, trust, and transparency is essential for building successful partnerships. Ethical dilemmas can propose challenges to researchers in global health. These challenges can be overcome by creating a shared vision for a research agenda, maintaining communication, and providing bidirectional training.

Deep multi-hybrid forecasting system with random EWT extraction and variational learning rate algorithm for crude oil futures

Machine learning algorithms provide feasibility for crude oil price prediction. In this paper, a novel multi-hybrid predictive neural network model is proposed based on complex deep learning algorithm, which integrates empirical wavelet transform, random inheritance formula error correction algorithm, deep bidirectional LSTM neural network and Elman recurrent neural network with variational learning rate. The prediction model is selected according to the sequence frequency after EWT feature extraction, and the prediction results are obtained by separately predicting and reintegrating. On the basis of individual model, the structure of deep bidirectional training, random inheritance formula and variational learning rate are proposed, which further ameliorate the performance of the model and achieve more effective data information capture. Simultaneously, the examination of variational learning rate provides us with a feasible parameter selection. The proposed model achieves high-precision prediction of crude oil futures price, and stands out in the multi-model comparison analysis and q-DSCID synchronous evaluation, with superior prediction accuracy.