A review of some techniques for inclusion of domain-knowledge into deep neural networks

Dash, Tirtharaj; Chitlangia, Sharad; Ahuja, Aditya; Srinivasan, Ashwin

doi:10.1038/s41598-021-04590-0

Cited by 95 publications

(46 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other proposals set out from deep learning architectures and add "logic" components to it. According to Dash et al [68], this can be done in a variety of ways, (a) by transforming data; (b) by transforming the loss function informed by a domain model; and (c) by transforming the model, e.g., by modeling logic operators within the network itself. All these methods appear generic enough to be applicable to movement analytics.…”

Section: Discussionmentioning

confidence: 99%

“…Logic can be a helpful mechanism to capture domain knowledge in deep learning architectures. Dash et al [68] Wan and Song [251] falls in category (a) and present an approach to add a set of auxiliary inputs to help interpret the outcome of a neural network. Input data is passed through a neural network to generate the auxiliary inputs to the next network.…”

Section: Logic and Deep Learningmentioning

confidence: 99%

See 1 more Smart Citation

Movement Analytics: Current Status, Application to Manufacturing, and Future Prospects from an AI Perspective

Baumgartner

Smith

Rana

et al. 2022

Preprint

View full text Add to dashboard Cite

Data-driven decision making is becoming an integral part of manufacturing companies. Data is collected and commonly used to improve efficiency and produce high quality items for the customers. IoT-based and other forms of object tracking are an emerging tool for collecting movement data of objects/entities (e.g. human workers, moving vehicles, trolleys etc.) over space and time. Movement data can provide valuable insights like process bottlenecks, resource utilization, effective working time etc. that can be used for decision making and improving efficiency. Turning movement data into valuable information for industrial management and decision making requires analysis methods. We refer to this process as movement analytics. The purpose of this document is to review the current state of work for movement analytics both in manufacturing and more broadly. We survey relevant work from both a theoretical perspective and an application perspective. From the theoretical perspective, we put an emphasis on useful methods from two research areas: machine learning, and logic-based knowledge representation. We also review their combinations in view of movement analytics, and we discuss promising areas for futuredevelopment and application. Furthermore, we touch on constraint optimization. From an application perspective, we review applications of these methods to movement analytics in a general sense and across various industries. We also describe currently available commercial off-the-shelf products for tracking in manufacturing, and we overview main concepts of digital twins and their applications.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Logic and Deep Learningmentioning

confidence: 99%

Movement Analytics: Current Status, Application to Manufacturing, and Future Prospects from an AI Perspective

Baumgartner

Smith

Rana

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In this section, we position SPLs against state-of-the-art approaches for enforcing constraints on neural network predictions. In-depth surveys on this topic can be found in [18] and [33].…”

Section: Related Workmentioning

confidence: 99%

“…Loss-based methods. A prominent strategy consists of penalizing the network for producing inconsistent predictions using an auxiliary loss [18,33]. While popular, loss-based methods, however cannot guarantee that the predictions will be consistent at test time.…”

Section: Related Workmentioning

confidence: 99%

Semantic Probabilistic Layers for Neuro-Symbolic Learning

Ahmed¹,

Teso²,

Chang³

et al. 2022

Preprint

View full text Add to dashboard Cite

We design a predictive layer for structured-output prediction (SOP) that can be plugged into any neural network guaranteeing its predictions are consistent with a set of predefined symbolic constraints. Our Semantic Probabilistic Layer (SPL) can model intricate correlations, and hard constraints, over a structured output space all while being amenable to end-to-end learning via maximum likelihood. SPLs combine exact probabilistic inference with logical reasoning in a clean and modular way, learning complex distributions and restricting their support to solutions of the constraint. As such, they can faithfully, and efficiently, model complex SOP tasks beyond the reach of alternative neuro-symbolic approaches. We empirically demonstrate that SPLs outperform these competitors in terms of accuracy on challenging SOP tasks including hierarchical multi-label classification, pathfinding and preference learning, while retaining perfect constraint satisfaction.Preprint. Under review.

show abstract

“…g A substantial growth in incorporating knowledge with deep learning has pursued. 3 Humans are capable of handling information at various levels of abstraction, which corresponds to different levels of learning. A knowledge-driven approach in AI should explore stratified knowledge manifested in various human-curated knowledge sources, generalpurpose or domain-specific, lexical or graph-based to find connections between facts and observations, yielding outcomes that humans can relate with their understanding and reason over the model.…”

mentioning

confidence: 99%