"Learning in reverse: Dopamine errors drive excitatory and inhibitory components of backward conditioning in an outcome-specific manner".

For over two decades, midbrain dopamine was considered synonymous with the prediction error in temporal-difference reinforcement learning. Central to this proposal is the notion that reward-predictive stimuli become endowed with the scalar value of predicted rewards. When these cues are subsequently encountered, their predictive value is compared to the value of the actual reward received allowing for the calculation of prediction errors. Phasic firing of dopamine neurons was proposed to reflect this computation, facilitating the backpropagation of value from the predicted reward to the reward-predictive stimulus, thus reducing future prediction errors. There are two critical assumptions of this proposal: 1) that dopamine errors can only facilitate learning about scalar value and not more complex features of predicted rewards, and 2) that the dopamine signal can only be involved in anticipatory learning in which cues or actions precede rewards. Recent work has challenged the first assumption, demonstrating that phasic dopamine signals across species are involved in learning about more complex features of the predicted outcomes, in a manner that transcends this value computation. Here, we tested the validity of the second assumption. Specifically, we examined whether phasic midbrain dopamine activity would be necessary for backward conditioning- when a neutral cue reliably follows a rewarding outcome. Using a specific Pavlovian-to-Instrumental Transfer (PIT) procedure, we show rats learn both excitatory and inhibitory components of a backward association, and that this association entails knowledge of the specific identity of the reward and cue. We demonstrate that brief optogenetic inhibition of ventral tegmental area dopamine (VTA DA) neurons timed to the transition between the reward and cue, reduces both of these components of backward conditioning. These findings suggest VTA DA neurons are capable of facilitating associations between contiguously occurring events, regardless of the content of those events. We conclude that these data are in line with suggestions that the VTA DA error acts as a universal teaching signal. This may provide insight into why dopamine function has been implicated in a myriad of psychological disorders that are characterized by very distinct reinforcement-learning deficits.

Improved Use of Drone Imagery for Malaria Vector Control through Technology-Assisted Digitizing (TAD)

Drones have the potential to revolutionize malaria vector control initiatives through rapid and accurate mapping of potential malarial mosquito larval habitats to help direct field Larval Source Management (LSM) efforts. However, there are no clear recommendations on how these habitats can be extracted from drone imagery in an operational context. This paper compares the results of two mapping approaches: supervised image classification using machine learning and Technology-Assisted Digitising (TAD) mapping that employs a new region growing tool suitable for non-experts. These approaches were applied concurrently to drone imagery acquired at seven sites in Zanzibar, United Republic of Tanzania. Whilst the two approaches were similar in processing time, the TAD approach significantly outperformed the supervised classification approach at all sites (t = 5.1, p < 0.01). Overall accuracy scores (mean overall accuracy 62%) suggest that a supervised classification approach is unsuitable for mapping potential malarial mosquito larval habitats in Zanzibar, whereas the TAD approach offers a simple and accurate (mean overall accuracy 96%) means of mapping these complex features. We recommend that this approach be used alongside targeted ground-based surveying (i.e., in areas inappropriate for drone surveying) for generating precise and accurate spatial intelligence to support operational LSM programmes.

Photocontrolled multiple-state photochromic benzo[b]phosphole thieno[3,2-b]phosphole-containing alkynylgold(I) complex via selective light irradiation

Photochromic materials have drawn growing attention because using light as a stimulus has been regarded as a convenient and environmental-friendly way to control properties of smart materials. While photoresponsive systems that are capable of showing multiple-state photochromism are attractive, the development of materials with such capabilities has remained a challenging task. Here we show that a benzo[b]phosphole thieno[3,2‑b]phosphole-containing alkynylgold(I) complex features multiple photoinduced color changes, in which the gold(I) metal center plays an important role in separating two photoactive units that leads to the suppression of intramolecular quenching processes of the excited states. More importantly, the exclusive photochemical reactivity of the thieno[3,2‑b]phosphole moiety of the gold(I) complex can be initiated upon photoirradiation of visible light. Stepwise photochromism of the gold(I) complex has been made possible, offering an effective strategy for the construction of multiple-state photochromic materials with multiple photocontrolled states to enhance the storage capacity of potential optical memory devices.

Maximizing Regulatory Review Efficiency: The Evolution of the FDA OCE RTOR Pilot

To promote the efficient review of oncology drug applications, the US Food and Drug Administration (FDA) Oncology Center of Excellence (OCE) launched the Real-Time Oncology Review (RTOR) pilot program in 2018. RTOR allows FDA to review individual sections of eCTD modules of a drug application for oncology drugs in contrast to requiring the applicant to submit complete modules or the complete application before review is initiated. Initially, the program accepted only supplemental applications with simple study designs and easily interpretable endpoints, but the scope has since been expanded to include applications for New Molecular Entities (NME), and other applications with more complex features. Though many applicants experience faster approvals under RTOR, it is difficult to isolate the effect of the RTOR program on review timelines as its contribution is masked by other expedited programs like priority review and breakthrough therapy designation (BTD). This article discusses the expanded scope of RTOR, its interplay with other OCE initiatives to modernize regulatory review, summarizes Genentech’s experiences in planning RTOR submissions from February 2019 to July 2021, and provides considerations for the future of the program.

Review of Machine Learning Applications Using Retinal Fundus Images

Automating screening and diagnosis in the medical field saves time and reduces the chances of misdiagnosis while saving on labor and cost for physicians. With the feasibility and development of deep learning methods, machines are now able to interpret complex features in medical data, which leads to rapid advancements in automation. Such efforts have been made in ophthalmology to analyze retinal images and build frameworks based on analysis for the identification of retinopathy and the assessment of its severity. This paper reviews recent state-of-the-art works utilizing the color fundus image taken from one of the imaging modalities used in ophthalmology. Specifically, the deep learning methods of automated screening and diagnosis for diabetic retinopathy (DR), age-related macular degeneration (AMD), and glaucoma are investigated. In addition, the machine learning techniques applied to the retinal vasculature extraction from the fundus image are covered. The challenges in developing these systems are also discussed.

Texture-like representation of objects in human visual cortex

The human visual ability to recognize objects and scenes is widely thought to rely on representations in category-selective regions of visual cortex. These representations could support object vision by specifically representing objects, or, more simply, by representing complex visual features regardless of the particular spatial arrangement needed to constitute real world objects. That is, by representing visual textures. To discriminate between these hypotheses, we leveraged an image synthesis approach that, unlike previous methods, provides independent control over the complexity and spatial arrangement of visual features. We found that human observers could easily detect a natural object among synthetic images with similar complex features that were spatially scrambled. However, observer models built from BOLD responses from category-selective regions, as well as a model of macaque inferotemporal cortex and Imagenet-trained deep convolutional neural networks, were all unable to identify the real object. This inability was not due to a lack of signal-to-noise, as all of these observer models could predict human performance in image categorization tasks. How then might these texture-like representations in category-selective regions support object perception? An image-specific readout from category-selective cortex yielded a representation that was more selective for natural feature arrangement, showing that the information necessary for object discrimination is available. Thus, our results suggest that the role of human category-selective visual cortex is not to explicitly encode objects but rather to provide a basis set of texture-like features that can be infinitely reconfigured to flexibly learn and identify new object categories.

Market Segmentation of Online Reviews: A Network Analysis Approach

With the popularity of online reviews, brand managers have opportunities to segment their markets according to the reviews of their products or services by customers. Nonetheless, it has been suggested that traditional market segmentation methods are ineffective at analyzing online review data due to the complex features and large amount of this type of data; specifically, traditional methods fail to take into account the networked nature of interactive relationships among reviewers and brands across online review websites. Accordingly, this study proposes a network analysis approach for the market segmentation of online reviews. Collecting samples from Yelp via web scraping, this study demonstrates how network analysis techniques can be utilized to segment online reviewers through a four-step process. The results reveal the core and peripheral market segments, as well as the bridge segment in the core. The study contributes to offering marketing researchers and managers a new network structure analysis approach for the market segmentation of online reviews.

Kernel Fisher Envelope Surface for Pattern Recognition

It is found that the batch process is more difficultly monitored compared with the continuous process, due to its complex features, such as nonlinearity, non-stable operation, unequal production cycles, and most variables only measured at the end of batch. Traditional methods for batch process, such as multiway FDA (Chen 2004) and multi-model FDA (He et al. 2005), cannot solve these issues well. They require complete batch data only available at the end of a batch. Therefore, the complete batch trajectory must be estimated real time, or alternatively only the measured values at the current moment are used for online diagnosis. Moreover, the above approaches do not consider the problem of inconsistent production cycles.

Case studies on seismic assessment of historical buildings using advanced analysis

Seismic risk mitigation in existing buildings requires an engineering assessment of the current condition and expected seismic performance and an identification of possible deficiencies that should be addressed. For heritage and historical buildings in particular, there is significant benefit in using the most detailed analysis methods available to avoid the conservatism inherent in simpler methods and thereby minimise unnecessary interventions and more precisely pinpoint where strengthening is required. On recent heritage projects, Arup has used the analysis software LS-DYNA and a new material model, calibrated against experimental tests on unreinforced masonry components and buildings to carry out (or supplement) seismic assessments. The analysis method (non-linear response history analysis) is not new, but its application on detailed finite-element models of complex historic structures has previously been computationally prohibitive and requires significant analyst experience to deliver reliable results. This paper summarises three of these recent Arup projects: Woltersum Church (Netherlands), Procuratie Vecchie (Venice) and a building cluster in the historical centre of Appingedam (Netherlands). The case studies show that these analyses allow complex features of seismic performance to be considered, such as damage or modifications to the building over time, pounding (separate buildings colliding into one another due to seismic movements) and load sharing between adjacent structures.

Using Vector Agents to Implement an Unsupervised Image Classification Algorithm

Unsupervised image classification methods conventionally use the spatial information of pixels to reduce the effect of speckled noise in the classified map. To extract this spatial information, they employ a predefined geometry, i.e., a fixed-size window or segmentation map. However, this coding of geometry lacks the necessary complexity to accurately reflect the spatial connectivity within objects in a scene. Additionally, there is no unique mathematical formula to determine the shape and scale applied to the geometry, being parameters that are usually estimated by expert users. In this paper, a novel geometry-led approach using Vector Agents (VAs) is proposed to address the above drawbacks in unsupervised classification algorithms. Our proposed method has two primary steps: (1) creating reliable training samples and (2) constructing the VA model. In the first step, the method applies the statistical information of a classified image by k-means to select a set of reliable training samples. Then, in the second step, the VAs are trained and constructed to classify the image. The model is tested for classification on three high spatial resolution images. The results show the enhanced capability of the VA model to reduce noise in images that have complex features, e.g., streets, buildings.