Perfect Recall Research Articles

Knowing How to Plan

Various planning-based know-how logics have been studied in the recent literature. In this paper, we use such a logic to do know-how-based planning via model checking. In particular, we can handle the higher-order epistemic planning involving know-how formulas as the goal, e.g., find a plan to make sure p such that the adversary does not know how to make p false in the future. We give a PTIME algorithm for the model checking problem over finite epistemic transition systems and axiomatize the logic under the assumption of perfect recall.

Dynamic Task Allocation in Multi-Robot System Based on a Team Competition Model.

In recent years, it is a trend to integrate the ideas in game theory into the research of multi-robot system. In this paper, a team-competition model is proposed to solve a dynamic multi-robot task allocation problem. The allocation problem asks how to assign tasks to robots such that the most suitable robot is selected to execute the most appropriate task, which arises in many real-life applications. To be specific, we study multi-round team competitions between two teams, where each team selects one of its players simultaneously in each round and each player can play at most once, which defines an extensive-form game with perfect recall. We also study a common variant where one team always selects its player before the other team in each round. Regarding the robots as the players in the first team and the tasks as the players in the second team, the sub-game perfect strategy of the first team computed via solving the team competition gives us a solution for allocating the tasks to the robots—it specifies how to select the robot (according to some probability distribution if the two teams move simultaneously) to execute the upcoming task in each round, based on the results of the matches in the previous rounds. Throughout this paper, many properties of the sub-game perfect equilibria of the team competition game are proved. We first show that uniformly random strategy is a sub-game perfect equilibrium strategy for both teams when there are no redundant players. Secondly, a team can safely abandon its weak players if it has redundant players and the strength of players is transitive. We then focus on the more interesting case where there are redundant players and the strength of players is not transitive. In this case, we obtain several counterintuitive results. For example, a player might help improve the payoff of its team, even if it is dominated by the entire other team. We also study the extent to which the dominated players can increase the payoff. Very similar results hold for the aforementioned variant where the two teams take actions in turn.

Funes the Crate Digger: AI, Perception, and Recall


 
 This essay is inspired by Jorge Luis Borges’s short story about Funes, a young man with perfect perception and recall. His stunning archival ability allowed Funes to create a shorthand system for creativity, a “rhapsody of unconnected words” to stand in for his archive of memories: “Anything he thought, even once, remained ineradicably with him.” There is similar malady today, a cultural hyperthymesia, a fetish for the archive. Information is recalcitrant and will not disappear, which cheapens the value of information while amplifying the need for human attention. The art of searching through the information rapids, then, depends upon elimination perhaps more than retrieval. Human curators with inclinations toward perception and recall are needed to filter through this influx of information rather than permitting the reliance upon algorithms by default. This essay theorizes the remix artist as a Borgesian Librarian, a curator of archived information fragments. Through the process of “crate digging,” remixers find long-forgotten samples of music and then exploit their allegorical potential. Like Funes, remix artists have an uncanny ability for perception and recall. Unlike Funes, rather than perceiving their talents as a curse, remix artists are model curators, exemplars of how to filter and make meaning from information overabundance. To illustrate these points, I will address the crate digging on display in two essential remix albums, DJ Shadow’s Endtroducing… and J Dilla’s Donuts.
 

Set-theory based benchmarking of three different variant callers for targeted sequencing

BackgroundNext generation sequencing (NGS) technologies have improved the study of hereditary diseases. Since the evaluation of bioinformatics pipelines is not straightforward, NGS demands effective strategies to analyze data that is of paramount relevance for decision making under a clinical scenario. According to the benchmarking framework of the Global Alliance for Genomics and Health (GA4GH), we implemented a new simple and user-friendly set-theory based method to assess variant callers using a gold standard variant set and high confidence regions. As model, we used TruSight Cardio kit sequencing data of the reference genome NA12878. This targeted sequencing kit is used to identify variants in key genes related to Inherited Cardiac Conditions (ICCs), a group of cardiovascular diseases with high rates of morbidity and mortality.ResultsWe implemented and compared three variant calling pipelines (Isaac, Freebayes, and VarScan). Performance metrics using our set-theory approach showed high-resolution pipelines and revealed: (1) a perfect recall of 1.000 for all three pipelines, (2) very high precision values, i.e. 0.987 for Freebayes, 0.928 for VarScan, and 1.000 for Isaac, when compared with the reference material, and (3) a ROC curve analysis with AUC > 0.94 for all cases. Moreover, significant differences were obtained between the three pipelines. In general, results indicate that the three pipelines were able to recognize the expected variants in the gold standard data set.ConclusionsOur set-theory approach to calculate metrics was able to identify the expected ICCs related variants by the three selected pipelines, but results were completely dependent on the algorithms. We emphasize the importance to assess pipelines using gold standard materials to achieve the most reliable results for clinical application.

Developing Software Signature Search Engines Using Paragraph Vector Model: A Triage Approach for Digital Forensics

Today, with the growth of information and communication technology, digital crimes have also spread. Advanced storage technologies and their low cost have led to a significant increase in their use. Therefore, the high volume of digital data to be analyzed is a challenge facing digital forensic investigators. Digital forensic triage solutions aim to alleviate the forensic backlog. A promising triage technique is to quickly find the software packages run on the target system to narrow down the search space. In this paper, we propose a software signature search engine (S3E) to identify software on the system under investigation. The document collection of this search engine consists of software signatures, and the query is the features extracted from the system’s hard disk. We propose a forensic differential analysis model to build software signatures. Besides, we use the paragraph vector model to construct the corresponding vectors of each software signature and find similarities between the query vector and the signature vectors. Different design parameters are involved in making software signature search engines, and distinct values of these parameters lead to different models. We have measured the performance of these S3E models against several controlled systems and some pseudo-real systems. The experimental results on both datasets show that some S3E models achieve perfect recall, and many of them have a recall of more than 90%. Besides, we find that the recall rate of the S3E models in both datasets is higher than the averaged word2vec model and the TF-IDF model.

PATH-22. COMPARISON OF SUPERVISED CLASSIFICATION METHODS FOR CENTRAL NERVOUS SYSTEM TUMORS BASED ON DNA-METHYLATION

Classification of brain tumors using methylation profiling is an important diagnostic advance, reducing subjectivity and improving interpretability of clinical outcome data. Despite the recognized value of methylation profiling in the clinical laboratory, the performance characteristics of different supervised classification models has not been directly compared. We developed 3 methods using methylation profiles to classify CNS tumors: an exact bootstrap k-nearest neighbor (kNN), a multi-layer perceptron neural net (NN), and a random forest classifier (RF). We trained these methods on the publicly available CNS tumor reference cohort (GSE90496) with 2,801 profiles and 91 classes. We evaluated the performance of these methods by leave-out-25% cross-validation. The relative performance of these methods were evaluated in terms of accuracy, precision, and recall for class or class family. The kNN, RF, and NN classifier had an estimate error rate of 10.74%, 4.01%, and 1.89%, respectively for class prediction and an error rate for family prediction of 5.97%, 0.90%, and 0.6%, respectively. At perfect recall for class assignment, the RF and kNN had a precision of 0.96 and 0.89 while the NN reached 0.98. For family assignment, the precision for the three classifiers was almost 1.0 with recall of nearly 0.8. At the recall rate of 1.0, the precision dropped to 0.94, 0.991 and 0.994 for kNN, RF, and NN, respectively. Overall, the NN showed improved performance metrics compared to the kNN and RF in CNS tumor classification for both class and class family assignment.

Undecidable Cases of Model Checking Probabilistic Temporal-Epistemic Logic

We investigate the decidability of model checking logics of time, knowledge, and probability, with respect to two epistemic semantics: the clock and synchronous perfect recall semantics in partially observable discrete-time Markov chains. Decidability results are known for certain restricted logics with respect to these semantics, subject to a variety of restrictions that are either unexplained or involve a longstanding unsolved mathematical problem. We show that mild generalizations of the known decidable cases suffice to render the model checking problem definitively undecidable. In particular, for the synchronous perfect recall semantics, a generalization from temporal operators with finite reach to operators with infinite reach renders model checking undecidable. The case of the clock semantics is closely related to a monadic second-order logic of time and probability that is known to be decidable, except on a set of measure zero. We show that two distinct extensions of this logic make model checking undecidable. One of these involves polynomial combinations of probability terms, the other involves monadic second-order quantification into the scope of probability operators. These results explain some of the restrictions in previous work.

Accuracy of Infection Control Surveillance in Identifying Genomically Confirmed Cross Transmission Clusters

Background: Infection prevention surveillance for cross transmission is often performed by manual review of microbiologic culture results to identify geotemporally related clusters. However, the sensitivity and specificity of this approach remains uncertain. Whole-genome sequencing (WGS) analysis can help provide a gold-standard for identifying cross-transmission events. Objective: We employed a published WGS program, the Philips IntelliSpace Epidemiology platform, to compare accuracy of two surveillance methods: (i.) a virtual infection practitioner (VIP) with perfect recall and automated analysis of antibiotic susceptibility testing (AST), sample collection timing, and patient location data and (ii) a novel clinical matching (CM) algorithm that provides cluster suggestions based on a nuanced weighted analysis of AST data, timing of sample collection, and shared location stays between patients. Methods: WGS was performed routinely on inpatient and emergency department isolates of Enterobacter cloacae, Enterococcus faecium, Klebsiella pneumoniae, and Pseudomonas aeruginosa at an academic medical center. Single-nucleotide variants (SNVs) were compared within core genome regions on a per-species basis to determine cross-transmission clusters. Moreover, one unique strain per patient was included within each analysis, and duplicates were excluded from the final results. Results: Between May 2018 and April 2019, clinical data from 121 patients were paired with WGS data from 28 E. cloacae, 21 E. faecium, 61 K. pneumoniae, and 46 P. aeruginosa isolates. Previously published SNV relatedness thresholds were applied to define genomically related isolates. Mapping of genomic relatedness defined clusters as follows: 4 patients in 2 E. faecium clusters and 2 patients in 1 P. aeruginosa cluster. The VIP method identified 12 potential clusters involving 28 patients, all of which were “pseudoclusters.” Importantly, the CM method identified 7 clusters consisting of 27 patients, which included 1 true E. faecium cluster of 2 patients with genomically related isolates. Conclusions: In light of the WGS data, all of the potential clusters identified by the VIP were pseudoclusters, lacking sufficient genomic relatedness. In contrast, the CM method showed increased sensitivity and specificity: it decreased the percentage of pseudoclusters by 14% and it identified a related genomic cluster of E. faecium. These findings suggest that integrating clinical data analytics and WGS is likely to benefit institutions in limiting expenditure of resources on pseudoclusters. Therefore, WGS combined with more sophisticated surveillance approaches, over standard methods as modeled by the VIP, are needed to better identify and address true cross-transmission events.Funding: This study was supported by Philips Healthcare.Disclosures: None

Extraction of Buildings in Urban Area for Surface Area Assessment from Satellite Imagery based on Morphological Building Index using SVM Classifier

Rooftops are essential features, extracted from satellite images for their significance in applications such as update of urban geodatabase, risk assessment and rescue map. In this work, a methodology (MBION-SVM) which integrates morphological, spectral, shape and geometrical features with SVM classifier to classify the objects within the satellite image into building rooftops and non-rooftops has been proposed. The probable buildings are detected using Morphological Building Index (MBI). The mislabeled rooftops are eliminated by combining Otsu thresholding and Normalized Difference Vegetation Index (NDVI). Geometrical features computed from identified building rooftops are used to train a support vector machine (SVM), and self-correction is performed for removing any mislabeled rooftops and to provide the data on surface area of the perfect rooftops. Here, we have used Very High Resolution (VHR) images of Worldview-2 and Sentinal-2. We have analyzed the performance of the proposed building extraction approach with classification algorithms such as linear discriminant analysis, logistic regression and SVM. Since the proposed method gives an accuracy around 99%, precision of 89%, a perfect recall of 1 and a F-score of 88%, it can be effectively utilized to extract the buildings from VHR images for any appropriate application.

Backward induction in games without perfect recall

The equilibrium concepts that we now think of as various forms of backwards induction, namely, subgame perfect equilibrium (Selten, 1965), perfect equilibrium (Selten, 1975), sequential equilibrium (Kreps and Wilson, 1982), and quasi-perfect equilibrium (van Damme, 1984), while formally well defined in a wider class of games, are explicitly restricted to games with perfect recall.We argue that in games without perfect recall the original definitions are inappropriate. Our reading of the original papers is that the authors were aware that their definitions did not formally require the assumption of perfect recall but they were also aware that without the assumption of perfect recall the definitions they gave were not the “correct” ones. We give definitions of these concepts that identify the same outcomes in games with perfect recall and behave well in linear games without perfect recall.

The complexity of computing a (quasi-)perfect equilibrium for an n-player extensive form game

We study the complexity of computing or approximating refinements of Nash equilibrium for finite n-player extensive form games of perfect recall (EFGPR), n≥3. Our results apply to a number of well-studied refinements, including sequential equilibrium, extensive-form perfect equilibrium, and quasi-perfect equilibrium.Informally, we show that, for all these refinements, approximating such a refined equilibrium for an n-player EFGPR is not any harder than (i.e., can be efficiently reduced to) approximating a Nash equilibrium for a 3-player normal form game. More specifically, we show that approximating such a refined equilibrium for a given EFGPR, within given desired precision, is FIXPa-complete. We also study corresponding notions of “almost” equilibrium for these refinements, and we show that computing one is PPAD-complete. (In all these cases our main results show containment in FIXPa and containment in PPAD. Hardness follows from earlier results for simpler games.)

The Modal Logic LEC for Changing Knowledge, Expressed in the Growing Language

We present the propositional logic LEC for the two epistemic modalities of current and stable knowledge used by an agent who system-atically enriches his language. A change in the linguistic resources of an agent as a result of certain cognitive processes is something that commonly happens. Our system is based on the logic LC intended to formalize the idea that the occurrence of changes induces the passage of time. Here, the primitive operator C read as: it changes that, defines the temporal succession of states of the world. The notion of current knowledge concerns variable components of the world and it may change over time. We represent it by the primitive operator k read as: the agent currently knows that, and assume that it has S5 properties. The second type of knowledge, symbolized by the primitive operator K read as: the agent stably knows that, relates to constant components of the world and it does not change. As a result of the axiomatic entanglement of C, K and k we show that stable knowledge satisfies axioms of S4.3. K and k modalities are not mutually definable, stable knowledge implies the current one and if the latter never changes, then it comes to be stable. The combination of K and k with the idea of an expanding language allows questioning of the so-called perfect recall principle. It cannot be maintained for both types of knowledge just because of changes in the vocabulary of the agent and possibly the growing spectrum of possible states of the world. We interpret LEC in the semantics of histories of epistemic changes and show that it is complete. Finally, we compare our logic with selected epistemic logics based on the concept of linear discrete time.

Generic Metadata Time Carving

Recovery of files can be a challenging task in file system investigations, and most carving techniques are based on file signatures or semantics within the file. However, these carving techniques often only recover the files, but not the metadata associated with the file. In this paper, we propose a novel, generic approach for carving metadata by searching for equal and co-located timestamps. The rationale is that there are some common metadata for files and directories within each file system. Our generic time carver provides potential timestamp locations for repeated timestamps in each metadata structure, identifying potential metadata for files. A semantic parser then filters the results with respect to the specific file system type. In our experiments, extraction of MFT entries in NTFS and inodes in Ext4 had near perfect precision for metadata entries with multiple equivalent timestamps, and for such metadata structures we obtained perfect recall for NTFS. For known file systems, we use the information found within identified metadata to recover files, and by recovering files and their associated metadata we increase the evidential value of recovered files.

Fault identification and severity analysis of rolling element bearings using phase space topology

This article presents the application of phase space topology and time-domain statistical features for rolling element bearing diagnostics in rotating machines under variable operating conditions. The results indicate very promising performance in identifying various faults with virtually perfect accuracy, recall, and precision. A comparison with the envelope analysis method is performed to show the superior performance of the proposed approach. In addition, the results demonstrate an outstanding prediction rate for the fault diameter of bearing defects.

Double-Oracle Sampling Method for Stackelberg Equilibrium Approximation in General-Sum Extensive-Form Games

The paper presents a new method for approximating Strong Stackelberg Equilibrium in general-sum sequential games with imperfect information and perfect recall. The proposed approach is generic as it does not rely on any specific properties of a particular game model. The method is based on iterative interleaving of the two following phases: (1) guided Monte Carlo Tree Search sampling of the Follower's strategy space and (2) building the Leader's behavior strategy tree for which the sampled Follower's strategy is an optimal response. The above solution scheme is evaluated with respect to expected Leader's utility and time requirements on three sets of interception games with variable characteristics, played on graphs. A comparison with three state-of-the-art MILP/LP-based methods shows that in vast majority of test cases proposed simulation-based approach leads to optimal Leader's strategies, while excelling the competitive methods in terms of better time scalability and lower memory requirements.

Timed ATL: Forget Memory, Just Count

In this paper we investigate the Timed Alternating-Time Temporal Logic (TATL), a discrete-time extension of ATL. In particular, we propose, systematize, and further study semantic variants of TATL, based on different notions of a strategy. The notions are derived from different assumptions about the agents’ memory and observational capabilities, and range from timed perfect recall to untimed memoryless plans. We also introduce a new semantics based on counting the number of visits to locations during the play. We show that all the semantics, except for the untimed memoryless one, are equivalent when punctuality constraints are not allowed in the formulae. In fact, abilities in all those notions of a strategy collapse to the “counting” semantics with only two actions allowed per location. On the other hand, this simple pattern does not extend to the full TATL.
 As a consequence, we establish a hierarchy of TATL semantics, based on the expressivity of the underlying strategies, and we show when some of the semantics coincide. In particular, we prove that more compact representations are possible for a reasonable subset of TATL specifications, which should improve the efficiency of model checking and strategy synthesis.

A differentiable homotopy method to compute perfect equilibria

The notion of perfect equilibrium was formulated by Selten (Int J Game Theory 4(1):25–55, 1975) as a strict refinement of Nash equilibrium. For an extensive-form game with perfect recall, every perfect equilibrium of its agent normal-form game yields a perfect equilibrium of the extensive-form game. This paper aims to develop a differentiable homotopy method for computing perfect equilibria of normal-form games. To accomplish this objective, we constitute an artificial game by introducing a continuously differentiable function of an extra variable. The artificial game defines a differentiable homotopy mapping and establishes the existence of a smooth path to a perfect equilibrium. For numerical comparison, we also describe a simplicial homotopy method. Numerical results show that the differentiable homotopy method is numerically stable and efficient and significantly outperforms the simplicial homotopy method especially when the problem is large.

"Memory compression" effects in visual working memory are contingent on explicit long-term memory.

Brady, Konkle, and Alvarez (2009) argued that statistical learning boosts the number of colors that can be held online in visual working memory (WM). They showed that when specific colors are consistently paired together in a WM task, subjects can take optimal advantage of these regularities to recall more colors, an effect they labeled memory compression. They proposed that memory compression is a product of visual statistical learning, an automatic apprehension of statistical regularities that has been shown in prior work to be disconnected from explicit learning. If statistical learning enables an expansion of the number of individuated representations in visual WM, it would require revision of virtually all models of capacity in this online memory system. That said, this provocative claim is inconsistent with multiple studies that have found no improvement in WM performance following numerous repetitions of specific sample displays (e.g., Logie, Brockmole, & Vandenbroucke, 2009; Olson & Jiang, 2004). Here, we replicate the Brady et al. (2009) findings but show that memory compression effects were restricted to subjects who had perfect explicit recall of the color pairs at the end of the study, suggesting that statistical regularities boosted performance by enabling contributions from long-term memory. Thus, while memory compression effects provide an interesting example of the tight collaboration between online and offline memory representations, they do not provide evidence that statistical regularities can augment the number of individuated representations that can be concurrently stored in visual WM. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

An Abstraction-Based Method for Verifying Strategic Properties in Multi-Agent Systems with Imperfect Information

We investigate the verification of Multi-agent Systems against strategic properties expressed in Alternating-time Temporal Logic under the assumptions of imperfect information and perfect recall. To this end, we develop a three-valued semantics for concurrent game structures upon which we define an abstraction method. We prove that concurrent game structures with imperfect information admit perfect information abstractions that preserve three-valued satisfaction. Further, we present a refinement procedure to deal with cases where the value of a specification is undefined. We illustrate the overall procedure in a variant of the Train Gate Controller scenario under imperfect information and perfect recall.

Sequential Equilibrium in Computational Games

We examine sequential equilibrium in the context of computational games (Halpern and Pass 2015), where agents are charged for computation. In such games, an agent can rationally choose to forget, so issues of imperfect recall arise. In this setting, we consider two notions of sequential equilibrium. One is an ex ante notion, where a player chooses his strategy before the game starts and is committed to it, but chooses it in such a way that it remains optimal even off the equilibrium path. The second is an interim notion, where a player can reconsider at each information set whether he is doing the “right” thing, and if not, can change his strategy. The two notions agree in games of perfect recall, but not in games of imperfect recall. Although the interim notion seems more appealing, in a companion article (Halpern and Pass 2016), we argue that there are some deep conceptual problems with it in standard games of imperfect recall. We show that the conceptual problems largely disappear in the computational setting. Moreover, in this setting, under natural assumptions, the two notions coincide.