Recursive Loop Research Articles

On the effects of recursive convolutional layers in convolutional neural networks

The Recursive Convolutional Layer (RCL) is a module that wraps a recursive feedback loop around a convolutional layer (CL). The RCL has been proposed to address some of the shortcomings of Convolutional Neural Networks (CNNs), as its unfolding increases the depth of a network without increasing the number of weights. We investigated the “naïve” substitution of CL with RCL on three base models: a 4-CL model, ResNet, DenseNet and their RCL-ized versions: C-FRPN, R-ResNet, and R-DenseNet using five image classification datasets. We find that this one-to-one replacement significantly improves the performances of the 4-CL model, but not those of ResNet or DenseNet. This led us to investigate the implication of the RCL substitution on the 4-CL model which reveals, among a number of properties, that RCLs are particularly efficient in shallow CNNs. We proceeded to re-visit the first set of experiments by gradually transforming the 4-CL model and the C-FRPN into respectively ResNet and R-ResNet, and find that the performance improvement is largely driven by the training regime whereas any depth increase negatively impacts the RCL-ized version. We conclude that the replacement of CLs by RCLs shows great potential in designing high-performance shallow CNNs.

Making virtual celebrity: Platformization and intermediation in digital cultural production

Focusing on intermediation in cultural production in a digital ecology consisting of multiple platforms mediating simultaneously converging and diverging industries, this study critically engages with the thesis that platforms act primarily as a disintermediation vector in a linear value chain within a single industry. With two empirical cases of intermediation in the emerging virtual celebrity sector – one organizing the recursive loop of prosumption, the other articulating authenticity against technological standardization and overproduction, the study shifts the focus away from questions of labor and agency of individual creators and unpacks the conditions of intermediation in the context of new industrial models of value creation, commodification, and division of labor. The empirical cases demonstrate that the implications of platforms for digital cultural production are paradoxical – while their business models lead to a centralized process of value capture, their flexible organizational forms may afford new distributed patterns of value creation.

Где и как возникают смыслы (материалы круглого стола)

The discussion aims to identify the interpretive mechanisms that provide an interface be­tween text and context (between language and the world, language and culture, language and society). It is a multi-level system of interfaces connected by inversion and recursion relations and operations. We identify this system as a pragmasemantics. It also acts as an interface platform for correlation/transformation of intra-system semantic units and extralinguistic objects within a set of possible worlds. At the same time, this platform is endowed with the autopoietic potential to generate new operating interfaces systems, as well as textualized structures of operations and interpretations (instructions, memory, reframing). The genera­tion of meaning is the result of the interaction of the system platform with the platform-as-context, due to the possibility that one generates and determines the operational potential of the other, acting a recursive loop or Möbius strip. The connection between subjectivity and meaningful semiosis is crucially important due to the key role of textualization and langua­ging. The signified are reproduced by signifiers and thus generate those processes that are structured by interactions and other practices. In complex systems (biological and beyond), the possibility of self-replication arises through recursive self-interpretation. Thus, subjectivi­ty acts as a kind of universal interface that determines the change of contexts, their interac­tion, the generation of possible new ones, ensuring the pro-creativity and pre-adaptability of the system (its stability and development.

Brain-imaging evidence for compression of binary sound sequences in human memory.

According to the language-of-thought hypothesis, regular sequences are compressed in human memory using recursive loops akin to a mental program that predicts future items. We tested this theory by probing memory for 16-item sequences made of two sounds. We recorded brain activity with functional MRI and magneto-encephalography (MEG) while participants listened to a hierarchy of sequences of variable complexity, whose minimal description required transition probabilities, chunking, or nested structures. Occasional deviant sounds probed the participants' knowledge of the sequence. We predicted that task difficulty and brain activity would be proportional to the complexity derived from the minimal description length in our formal language. Furthermore, activity should increase with complexity for learned sequences, and decrease with complexity for deviants. These predictions were upheld in both fMRI and MEG, indicating that sequence predictions are highly dependent on sequence structure and become weaker and delayed as complexity increases. The proposed language recruited bilateral superior temporal, precentral, anterior intraparietal, and cerebellar cortices. These regions overlapped extensively with a localizer for mathematical calculation, and much less with spoken or written language processing. We propose that these areas collectively encode regular sequences as repetitions with variations and their recursive composition into nested structures.

TOTEs and trauma 2: a rat maternal care and stress epigenetic schematic

PurposeThis conceptual article presents a schematic of rat maternal behavior and niche stress epigenetic effects as a case study that is then aligned with current evolutionary concepts, which raises new questions regarding immigrant assimilation and niche dynamics.Design/methodology/approachThe necessary background material for rat maternal and niche(s) stress factors are incorporated into a recursive, test-operate-test (rTOT), information-only-transfer, schematic (Patterson, 2023), which is an extension and refinement of the test-operate-test-exit (TOTE) schematic of Miller et al. (1960).FindingsThe generated epigenetic rTOT demonstrates the fundamental evolutionary unit of the flexible organism within its niche(s). The rTOT also confirms that epigenetic processes, epigenetic inheritance and phenotype plasticity are significant conceptual tools for understanding evolution. The teleology of rat adaptations for niche fitness via maternal behavior has been demonstrated. Sterling's (2011) allostasis, or predictive homeostasis, is extended to include species-niche(s) interaction(s) that are governed by recursive information feedback loops that function via self-organized criticality (SOC) for species and niche(s). Use of a rat model for biosocial issues in humans is strengthened.Research limitations/implicationsEpigenetic rTOT only covers the species side of the evolutionary unit. Niche(s) require(s) a separate rTOT schematic. The information modeled does not include the entire system producing epigenetic effects but models a substantial portion of it.Practical implicationsEpigenetic rTOT demonstrates the utility of phenotypic plasticity, epigenetics and epigenetic inheritance as explanations for inheritable behavior patterns. rTOT is a useful computational model for evolutionary issues. The issues involved in niche modeling using an rTOT schematic are briefly reviewed.Social implicationsWhen the demonstrated epigenetic model of rat genetics and inherited behavior are applied to the issues of immigrant enclaves, epigenetic complications for the difficulties of assimilation into the culture within which the enclaves are embedded become apparent. However, the questions raised must be addressed with extreme care to avoid cultural imperialism. Such cultural issues must be modeled with an rTOT application that covers the materials involved. The limitations of human Learning III restrictions when attempting to model Learning IV issues are addressed. Research into the means by which abuse and trauma are maintained by epigenetic means is urgently needed.Originality/valueThe rTOT schematic visualizes rat maternal behavior and stress epigenetic effects that produce inheritable behavior patterns, which answers Jablonka's (2017) request for new computational modeling representations. The concept of allostasis, or predictive homeostasis, (Sterling, 2011) is extended to the niche(s) of the organism under study so that allostasis becomes a fully cybernetic concept governed by SOC for both the organism and its niche(s). This new case study confirmed evolutionary effects of epigenetics, epigenetic inheritance and phenotypic plasticity. Niche control of organism evolution is presented. Epigenetic applications for immigrant assimilation issues have been suggested and niche dynamic questions have been raised.

Inverting resolution: accounting for the planetary cost of earth observation

This article uses the resolutional relationship between digital image and planetary surface in satellite remote sensing as a lens through which to view the reliance of visual culture on mineral resources. While most studies of resolution in satellite imaging have focused on visibility and invisibility, the author argues that the equivalence between pictorial and geographic space in its cm/pixel specification offers an opportunity to consider the physical relationship between the two. The proposed inversion enables the satellite and its transmitted images to be understood as contingent upon an unsustainably extractive industrial model. The article then traces the material trajectory in geophysical prospecting applications of remote sensing, identifying a recursive loop in which images are used to produce minerals that are used to produce images. The potential geopolitical impact of this circularity is then assessed with regard to an example of future remote sensing emissions governance. In this, and many other climate-critical applications, remote sensing potentially plays a vital role, yet its instrumental gaze tends to shape the earth as an informational resource whose mineral reserves should be capitalized upon. Ultimately, the author’s aim is not to denounce earth observation as ecologically untenable, but to propose that we find a measure with which to assess the planetary impact of the various aspects of industrialized visual culture. Conceiving of resolution not as a measure of pictorial space but of the terrestrial cost of producing, consuming, distributing and permanently storing digital images might enable a more relational understanding of image and ground, pixels and planet. Accounting for the inverse resolution of an image could bring the deep temporal costs of digital visual culture into focus.

Everybody Wants to be a Fascist Online: Psychoanalysis and the Digital Architecture of Fascism

Félix Guattari and Franco Berardi have both argued that media ecologies and psychic ecologies are intimately intertwined and as such, any exploration of the collective unconscious must engage with how the mind is formed with and through media. This understanding of networks of interdependence necessitates an exploration of how platformization has impacted users” collective psyche. Drawing from psychosocial theory, psychoanalysis and the work of Félix Guattari, this article analizes the micropolitics of desire of digital platforms, with an explicit focus on how algorithmic structures amplify extreme Right content, allowing fascisms to metastasis throughout digital spaces. It will first examine the algorithmic architecture of social media platforms, demonstrating how these digital spaces lock in and over-code desire through recursive feedback loops that amplify extremism. Following this will be an exploration of the excess of desire that is cut off and left as a remainder partial object, termed the “fascist abject,” and what role this process plays in the production of subjectivity.

Single-timepoint low-dimensional characterization and classification of acute versus chronic multiple sclerosis lesions using machine learning

Multiple sclerosis (MS) is a chronic inflammatory and neurodegenerative disease characterized by the appearance of focal lesions across the central nervous system. The discrimination of acute from chronic MS lesions may yield novel biomarkers of inflammatory disease activity which may support patient management in the clinical setting and provide endpoints in clinical trials. On a single timepoint and in the absence of a prior reference scan, existing methods for acute lesion detection rely on the segmentation of hyperintense foci on post-gadolinium T1-weighted magnetic resonance imaging (MRI), which may underestimate recent acute lesion activity. In this paper, we aim to improve the sensitivity of acute MS lesion detection in the single-timepoint setting, by developing a novel machine learning approach for the automatic detection of acute MS lesions, using single-timepoint conventional non-contrast T1- and T2-weighted brain MRI. The MRI input data are supplemented via the use of a convolutional neural network generating “lesion-free” reconstructions from original “lesion-present” scans using image inpainting. A multi-objective statistical ranking module evaluates the relevance of textural radiomic features from the core and periphery of lesion sites, compared within “lesion-free” versus “lesion-present” image pairs. Then, an ensemble classifier is optimized through a recursive loop seeking consensus both in the feature space (via a greedy feature-pruning approach) and in the classifier space (via model selection repeated after each pruning operation). This leads to the identification of a compact textural signature characterizing lesion phenotype. On the patch-level task of acute versus chronic MS lesion classification, our method achieves a balanced accuracy in the range of 74.3–74.6% on fully external validation cohorts.

FACCU: Enable Fast Accumulation for High-Speed DSP Systems

A number of fast accumulation (FACCU) designs are proposed for one of the essential components in DSP systems, the accumulator, to largely boost the DSP’s maximum attainable processing speed. By leveraging the unique features of the accumulator, the loop-incorporated critical path is reduced to as short as one or two full adders without degrading the sampling rate, thus efficiently addressing the system bottleneck caused by the recursive loop in accumulation. Based on this innovation, several highly scalable FACCU architectures are developed to meet various high-speed requirements with minimal overhead. Moreover, novel system-level co-optimization skills are developed in FACCU variants for two commonly used subsystems involving accumulators (the AdderTree-SA and the Distributed Arithmetic architecture) to achieve the best trade-off between area and speed, which have not been well investigated in previous works. Implementation results under the SMIC 55nm technology demonstrate that the proposed FACCU designs have great superiority in critical path delay, area, and latency compared with both prior art and the accumulators deeply optimized with the EDA tool. The application-level evaluations are also conducted for an FACCU-incorporated hardware accelerator of image processing network, demonstrating that the developed FACCU can efficiently improve the throughput in a real system.

A Unified Strategy for Formal Derivation and Proof of Binary Tree Nonrecursive Algorithms

In the formal derivation and proof of binary tree algorithms, Dijkstra's weakest predicate method is commonly used. However, the method has some drawbacks, including a time-consuming derivation process, complicated loop invariants, and the inability to generate executable programs from the specification. This paper proposes a unified strategy for the formal derivation and proof of binary tree non-recursive algorithms to address these issues. First, binary tree problem solving sequences are decomposed into two types of recursive relations based on queue and stack, and two corresponding loop invariant templates are constructed. Second, high-reliability Apla (abstract programming language) programs are derived using recursive relations and loop invariants. Finally, Apla programs are converted automatically into C++ executable programs. Two types of problems with binary tree queue and stack recursive relations are used as examples, and their formal derivation and proof are performed to validate the proposed strategy's effectiveness. This strategy improves the efficiency and correctness of binary tree algorithm derivation.

An Efficient Sorting Algorithm for Capacitor Voltage Balance of Modular Multilevel Converter With Space Vector Pulsewidth Modulation

Thisarticle presents an efficient analogue sorting algorithm for balancing the submodule (SM) capacitor voltages of modular multilevel converter (MMC). The proposed analogue sorting algorithm offers the advantage of fast convergence rate without any need of recursive loops for the implementation on embedded devices. It can be easily implemented with combinational logic operations on field programmable gate array (FPGA) and provides less hardware and computational overhead. The functionality and performance of the proposed analogue sorting algorithm is evaluated with the simulation model of three phase five-level MMC in MATLAB/Simulink environment. The real time implementation of the proposed sorting algorithm with the SM capacitor voltage balancing strategy is implemented on Altera/Cylone - I (EP1C12Q240C8N) FPGA. A five-level continuous space vector pulsewidth modulation (CSVPWM) is realized on a PIC microcontroller (PIC18F452). A down-scaled model of single-phase five-level MMC is designed and constructed to investigate the reliable and stable operation of MMC with the proposed analogue sorting algorithm and SVPWM method. Simulation and experimental results are presented for validation.

The Event/Machine of Neural Machine Translation?

ABSTRACT The present paper purports to question the status of machine translation as a technique, or even as a technology. The use of new technologies, and more particularly neural machine translation is problematic, because the operating mode of machine translation software is a machinic language processing that relies on the harvesting of potentially infinite databases to feed an artificial intelligence. Such a recursive loop only reinforces the sphere of the same without ever taking into account the risk of radical otherness in all its singularity. This is the spawn of our collective intelligence fed back to us by a network of interconnected algorithms in an endless feedback loop. I contend that translation allows us to overcome the dichotomy between techne and poiesis by opening the space for the emergence of otherness in the translating language that welcomes it. Even though the automaticity of the machine is antinomic with any poetic event, (how) can we, as Derrida puts it, “think together the machine and the event, a machinelike repetition and that which happens/arrives” 1 through (machine) translation?

Innovation intermediaries as collaborators in shaping service ecosystems: The importance of dynamic capabilities

Innovation is increasingly a cocreative process centered on developing new solutions and (re)shaping service ecosystems. Recent research provides important insights into the capabilities keystone actors deploy to drive innovation in their role as a “shaper”. These studies highlight other non-focal actors as supporters of innovation, but their roles and capabilities in shaping service ecosystems are less clear. To address this gap, we study innovation intermediaries in the shaping of a service ecosystem that encompasses many service providers beyond a central keystone actor. Using data from multiple cases within a healthcare service ecosystem, we show how innovation intermediaries make connections among diverse organizations as both shapers and supporters of service-ecosystem shaping. Specifically, innovation intermediaries shape ecosystem configuration by organizing resources and relationships and support ecosystem governance by adapting institutional arrangements. Our data reveal that the integration of these shaper and supporter roles scaffold the enactment of a complex role of collaborator. This role is filtered through a recursive design feedback loop of reflexivity and reformation, and grounded in dynamic capabilities of sensing, seizing, and reconfiguring. We find that the collaborator role is critical for co-designing joint value propositions and advancing the efforts of keystone actors to shape service ecosystems.

Making the Miscellany: Poetry, Print, and the History of the Book in Early Modern England

Making the Miscellany: Poetry, Print, and the History of the Book in Early Modern England

A Method to Deduce and Synthesize the Dafny Programs

We propose a systematic method to deduce and synthesize the Dafny programs. First, the specification of problem is described in strict mathematical language. Then, the derivation process uses program specification transformation technology to perform equivalent transformation. Furthermore, Dafny program is synthesized through the obtained recursive relationship and loop invariants. Finally, the functional correctness of Dafny program is automatically verified by Dafny verifier or online tool. Through this method, we deduce and synthesize Dafny programs for many typical problems such as the cube sum problem, the minimum (or maximum) contiguous subarray problems, several searching problems, several sorting problems, and so on. Due to space limitation, we only illustrate the development process of Dafny programs for two typical problems: the minimum contiguous subarray problem and the new local bubble sorting problem. It proves that our method can effectively improve the correctness and reliability of Dafny program developed. What’s more, we demonstrate the potential of the deductive synthesis method by developing a new local bubble Sorting program.

Futures literacy in the loop

This article proposes a functional historicist explanation to explicate the core ideas and underlying logic embedded in the futures literacy concept.Futures literacy assumes a capacity to reflect on the past, sense and make sense of the present and use this reflective body of knowledge when anticipating the future.Arguably, futures literacy must be learned, sustained, and regained; it requires a continuous, anticipative, and recursive loop. Recursivity, where an effect in an initial period acts as a cause in the next period, retroacts between the future and present, regaining anticipation. Anticipation has causal effects in the way it structures our images of the future and the avenue we follow when striving to achieve this image. Such a causal structure implies both feedforward and feedback control and is contained in the logic of functional explanations used in sociology.

Transformative or Tokenistic?

Transformative or Tokenistic?

A digital servitization framework for viable manufacturing companies

Purpose The purpose of this paper is to understand the strategic management of a technology-enabled shift from a product-centric to a service-centric logic and to identify the sociotechnical dynamics underlying this transition. The study focuses on how manufacturers manage to create value in industrial markets through digital servitization. Design/methodology/approach An abductive research approach is used to investigate two manufacturing firms, and an interpretive framework is used as an analytical template. A cross-case analysis is conducted. Findings The case companies strategically managed sociotechnical processes of digitization to co-create value. Their service orientation delineates dissimilarity in terms of digital servitization. It reflects a viable ecosystem that moves toward datatization through adaptation in one case and a viable ecosystem that moves toward digitization through reconfiguration in the other case. Practical implications A theoretically grounded, empirically informed framework is proposed to detect transformational mechanisms to manage value co-creation in digitally servitized contexts, thus contributing to ecosystem viability. Originality/value This is the first study to adopt a system perspective such as the viable system approach combined with service-dominant logic to reconceptualize the overall sociotechnical processes and the underlying mechanisms leading to digitized value creation. In line with a systems view and a systematic process based on a transformative attitude toward digital servitization, the empirically informed framework identifies specific co-creation activities and recursive feedback loops.

Simple Behavioral Model of Baseband Pulse Devices in the Form of a Second-Order Nonlinear Recursive Filter

A minimalistic nonlinear dynamic model, which still reflects the main aspects of the devices response to baseband pulse signals, is considered. The model is built as “black-box” and does not require knowing the structure and physical properties of the device. Since the model should reflect the exponential-type transient response, it is based on a nonlinear recursive filter. In the recursive loops the integrators are placed, because the main effect in the response of the simulated devices is the smoothing of the signal edge. The model reflects the nonlinear amplitude characteristic of the device (at the flat top of pulse), the dependence of the transient process duration on the input signal, and the nonlinear properties of the overshoot at the flat top of the pulse. We consider the realization of this model in CAD system and the methods for extracting the model parameters. The simulation error with the proposed model is 2-10% depending on the similarity of the test signal and signal at which the parameters were extracted.

The Power of Assembly

The Power of Assembly