Search

A microscopic model of a heterostructure with a quantum well (QW) is proposed to study the exciton behavior in an external electric field. The effect of an electric field ranging from 0 to 6 kV/cm applied to the GaAs/AlGaAs QW structure in the growth direction is studied for several QWs of various widths up to 100 nm. The three-dimensional Schr\"odinger equation (SE) of exciton is numerically solved using the finite difference method. Wave functions and energies for several states of the heavy-hole and light-hole excitons are calculated. Dependencies of the exciton state energy, the binding energy, the radiative broadening, and the static dipole moment on the applied electric fields are determined. The threshold of exciton dissociation for the 100-nm QW is also determined. In addition, we found the electric-field-induced shift of the center of mass of the heavy-hole and light-hole exciton in the QWs. Finally, we have modeled reflection spectra of heterostructures with the GaAs/AlGaAs QWs in the electric field using the calculated energies and radiative broadenings of excitons., Comment: 12 pages, 8 figures, to be published in Physical Review B

Despite the practical success of Artificial Intelligence (AI), current neural AI algorithms face two significant issues. First, the decisions made by neural architectures are often prone to bias and brittleness. Second, when a chain of reasoning is required, neural systems often perform poorly. Neuro-symbolic artificial intelligence is a promising approach that tackles these (and other) weaknesses by combining the power of neural perception and symbolic reasoning. Meanwhile, the success of AI has made it critical to understand its behaviour, leading to the development of explainable artificial intelligence (XAI). While neuro-symbolic AI systems have important advantages over purely neural AI, we still need to explain their actions, which are obscured by the interactions of the neural and symbolic components. To address the issue, this paper proposes a formal approach to explaining the decisions of neuro-symbolic systems. The approach hinges on the use of formal abductive explanations and on solving the neuro-symbolic explainability problem hierarchically. Namely, it first computes a formal explanation for the symbolic component of the system, which serves to identify a subset of the individual parts of neural information that needs to be explained. This is followed by explaining only those individual neural inputs, independently of each other, which facilitates succinctness of hierarchical formal explanations and helps to increase the overall performance of the approach. Experimental results for a few complex reasoning tasks demonstrate practical efficiency of the proposed approach, in comparison to purely neural systems, from the perspective of explanation size, explanation time, training time, model sizes, and the quality of explanations reported.

This paper addresses the problem of autonomous UAV search missions, where a UAV must locate specific Entities of Interest (EOIs) within a time limit, based on brief descriptions in large, hazard-prone environments with keep-out zones. The UAV must perceive, reason, and make decisions with limited and uncertain information. We propose NEUSIS, a compositional neuro-symbolic system designed for interpretable UAV search and navigation in realistic scenarios. NEUSIS integrates neuro-symbolic visual perception, reasoning, and grounding (GRiD) to process raw sensory inputs, maintains a probabilistic world model for environment representation, and uses a hierarchical planning component (SNaC) for efficient path planning. Experimental results from simulated urban search missions using AirSim and Unreal Engine show that NEUSIS outperforms a state-of-the-art (SOTA) vision-language model and a SOTA search planning model in success rate, search efficiency, and 3D localization. These results demonstrate the effectiveness of our compositional neuro-symbolic approach in handling complex, real-world scenarios, making it a promising solution for autonomous UAV systems in search missions.

The present bias is a well-documented behavioral trait that significantly influences human decision-making, with present-biased agents often prioritizing immediate rewards over long-term benefits, leading to suboptimal outcomes in various real-world scenarios. Kleinberg and Oren (2014) proposed a popular graph-theoretical model of inconsistent planning to capture the behavior of present-biased agents. In this model, a multi-step project is represented by a weighted directed acyclic task graph, where the agent traverses the graph based on present-biased preferences. We use the model of Kleinberg and Oren to address the principal-agent problem, where a principal, fully aware of the agent's present bias, aims to modify an existing project by adding or deleting tasks. The challenge is to create a modified project that satisfies two somewhat contradictory conditions. On one hand, the present-biased agent should select specific tasks deemed important by the principal. On the other hand, if the anticipated costs in the modified project become too high for the agent, there is a risk of the agent abandoning the entire project, which is not in the principal's interest. To tackle this issue, we leverage the tools of parameterized complexity to investigate whether the principal's strategy can be efficiently identified. We provide algorithms and complexity bounds for this problem., Comment: Accepted at ECAI 2024

Mesospheric CO2 clouds are one of two types of carbon dioxide clouds known on Mars. We present observations of mesospheric CO2 clouds made by Atmospheric Chemistry Suit (ACS) onboard the ESA-Roscosmos ExoMars Trace Gas Orbiter (TGO). We analyzed 1663 solar occultation sessions of Thermal InfraRed (TIRVIM) and Middle InfraRed (MIR) channels of ACS covering more than two Martian years that contain spectra of 2.7 {\mu}m carbon dioxide ice absorption band. That allowed us to unambiguously discriminate carbon dioxide ice aerosols from mineral dust and water ice aerosols, not relying on the information of atmospheric thermal conditions. CO2 clouds were detected in eleven solar occultation observations at altitudes from 39 km to 90 km. In five cases, there were two or three layers of CO2 clouds that were vertically separated by 5-15 km gaps. Effective radius of CO2 aerosol particles is in the range of 0.1-2.2 {\mu}m. Spectra produced by the smallest particles indicate a need for a better resolved CO2 ice refractive index. Nadir optical depth of CO2 clouds is in the range 5*10^{-4}-4*10^{-2} at both 2.7 {\mu}m and 0.8 {\mu}m. Asymmetrical diurnal distribution of detections observed by ACS is potentially due to local time variations of temperature induced by thermal tides. Two out of five cases of carbon dioxide cloud detections made by the TIRVIM instrument reveal the simultaneous presence of CO2 ice and H2O ice aerosols. Temperature profiles measured by the Near InfraRed (NIR) channel of ACS are used to calculate CO2 saturation ratio S at locations of carbon dioxide clouds. Supersaturation S > 1 is detected in only 5 out of 19 cases of CO2 cloud layers; extremely low values of S < 0.1 are found in 9 out of 19 cases., Comment: 29 pages, 10 figures. Submitted to Icarus

The uses of machine learning (ML) have snowballed in recent years. In many cases, ML models are highly complex, and their operation is beyond the understanding of human decision-makers. Nevertheless, some uses of ML models involve high-stakes and safety-critical applications. Explainable artificial intelligence (XAI) aims to help human decision-makers in understanding the operation of such complex ML models, thus eliciting trust in their operation. Unfortunately, the majority of past XAI work is based on informal approaches, that offer no guarantees of rigor. Unsurprisingly, there exists comprehensive experimental and theoretical evidence confirming that informal methods of XAI can provide human-decision makers with erroneous information. Logic-based XAI represents a rigorous approach to explainability; it is model-based and offers the strongest guarantees of rigor of computed explanations. However, a well-known drawback of logic-based XAI is the complexity of logic reasoning, especially for highly complex ML models. Recent work proposed distance-restricted explanations, i.e. explanations that are rigorous provided the distance to a given input is small enough. Distance-restricted explainability is tightly related with adversarial robustness, and it has been shown to scale for moderately complex ML models, but the number of inputs still represents a key limiting factor. This paper investigates novel algorithms for scaling up the performance of logic-based explainers when computing and enumerating ML model explanations with a large number of inputs.

FarView is an early-stage concept for a large, low-frequency radio observatory, manufactured in-situ on the lunar far side using metals extracted from the lunar regolith. It consists of 100,000 dipole antennas in compact subarrays distributed over a large area but with empty space between subarrays in a core-halo structure. FarView covers a total area of ~200 km2, has a dense core within the inner ~36 km2, and a ~power-law falloff of antenna density out to ~14 km from the center. With this design, it is relatively easy to identify multiple viable build sites on the lunar far side. The science case for FarView emphasizes the unique capabilities to probe the unexplored Cosmic Dark Ages - identified by the 2020 Astrophysics Decadal Survey as the discovery area for cosmology. FarView will deliver power spectra and tomographic maps tracing the evolution of the Universe from before the birth of the first stars to the beginning of Cosmic Dawn, and potentially provide unique insights into dark matter, early dark energy, neutrino masses, and the physics of inflation. What makes FarView feasible and affordable in the timeframe of the 2030s is that it is manufactured in-situ, utilizing space industrial technologies. This in-situ manufacturing architecture utilizes Earth-built equipment that is transported to the lunar surface to extract metals from the regolith and will use those metals to manufacture most of the array components: dipole antennas, power lines, and silicon solar cell power systems. This approach also enables a long functional lifetime, by permitting servicing and repair of the observatory. The full 100,000 dipole FarView observatory will take 4 - 8 years to build, depending on the realized performance of the manufacturing elements and the lunar delivery scenario., Comment: 26 pages, 7 figures, 2 tables

Widespread use of artificial intelligence (AI) algorithms and machine learning (ML) models on the one hand and a number of crucial issues pertaining to them warrant the need for explainable artificial intelligence (XAI). A key explainability question is: given this decision was made, what are the input features which contributed to the decision? Although a range of XAI approaches exist to tackle this problem, most of them have significant limitations. Heuristic XAI approaches suffer from the lack of quality guarantees, and often try to approximate Shapley values, which is not the same as explaining which features contribute to a decision. A recent alternative is so-called formal feature attribution (FFA), which defines feature importance as the fraction of formal abductive explanations (AXp's) containing the given feature. This measures feature importance from the view of formally reasoning about the model's behavior. It is challenging to compute FFA using its definition because that involves counting AXp's, although one can approximate it. Based on these results, this paper makes several contributions. First, it gives compelling evidence that computing FFA is intractable, even if the set of contrastive formal explanations (CXp's) is provided, by proving that the problem is #P-hard. Second, by using the duality between AXp's and CXp's, it proposes an efficient heuristic to switch from CXp enumeration to AXp enumeration on-the-fly resulting in an adaptive explanation enumeration algorithm effectively approximating FFA in an anytime fashion. Finally, experimental results obtained on a range of widely used datasets demonstrate the effectiveness of the proposed FFA approximation approach in terms of the error of FFA approximation as well as the number of explanations computed and their diversity given a fixed time limit.

Recent years have witnessed the widespread use of artificial intelligence (AI) algorithms and machine learning (ML) models. Despite their tremendous success, a number of vital problems like ML model brittleness, their fairness, and the lack of interpretability warrant the need for the active developments in explainable artificial intelligence (XAI) and formal ML model verification. The two major lines of work in XAI include feature selection methods, e.g. Anchors, and feature attribution techniques, e.g. LIME and SHAP. Despite their promise, most of the existing feature selection and attribution approaches are susceptible to a range of critical issues, including explanation unsoundness and out-of-distribution sampling. A recent formal approach to XAI (FXAI) although serving as an alternative to the above and free of these issues suffers from a few other limitations. For instance and besides the scalability limitation, the formal approach is unable to tackle the feature attribution problem. Additionally, a formal explanation despite being formally sound is typically quite large, which hampers its applicability in practical settings. Motivated by the above, this paper proposes a way to apply the apparatus of formal XAI to the case of feature attribution based on formal explanation enumeration. Formal feature attribution (FFA) is argued to be advantageous over the existing methods, both formal and non-formal. Given the practical complexity of the problem, the paper then proposes an efficient technique for approximating exact FFA. Finally, it offers experimental evidence of the effectiveness of the proposed approximate FFA in comparison to the existing feature attribution algorithms not only in terms of feature importance and but also in terms of their relative order.

In the quest for Explainable Artificial Intelligence (XAI) one of the questions that frequently arises given a decision made by an AI system is, ``why was the decision made in this way?'' Formal approaches to explainability build a formal model of the AI system and use this to reason about the properties of the system. Given a set of feature values for an instance to be explained, and a resulting decision, a formal abductive explanation is a set of features, such that if they take the given value will always lead to the same decision. This explanation is useful, it shows that only some features were used in making the final decision. But it is narrow, it only shows that if the selected features take their given values the decision is unchanged. It's possible that some features may change values and still lead to the same decision. In this paper we formally define inflated explanations which is a set of features, and for each feature of set of values (always including the value of the instance being explained), such that the decision will remain unchanged. Inflated explanations are more informative than abductive explanations since e.g they allow us to see if the exact value of a feature is important, or it could be any nearby value. Overall they allow us to better understand the role of each feature in the decision. We show that we can compute inflated explanations for not that much greater cost than abductive explanations, and that we can extend duality results for abductive explanations also to inflated explanations.

Exciton energy structure and population dynamics in a wide CdTe/CdZnTe quantum well are studied by spectrally-resolved pump-probe spectroscopy. Multiple excitonic resonances in reflectance spectra are observed and identified by solving numerically three-dimensional Schr\"odinger equation. The pump-probe reflectivity signal is shown to be dominated by the photoinduced nonradiative broadening of the excitonic resonances, while pump-induced exciton energy shift and reduction of the oscillator strength appear to be negligible. This broadening is induced by the reservoir of dark excitons with large in-plane wave vector, which are coupled to the the bright excitons states. The dynamics of the pump-induced nonradiative broadening observed experimentally is characterised by three components: signal build up on the scale of tens of picoseconds (i) and bi-exponential decay on the scale of one nanosecond (ii) and ten nanosecons (iii). Possible mechanisms of the reservoir population and depletion responsible for this behaviour are discussed., Comment: 8 pages, 3 figures

The most widely studied explainable AI (XAI) approaches are unsound. This is the case with well-known model-agnostic explanation approaches, and it is also the case with approaches based on saliency maps. One solution is to consider intrinsic interpretability, which does not exhibit the drawback of unsoundness. Unfortunately, intrinsic interpretability can display unwieldy explanation redundancy. Formal explainability represents the alternative to these non-rigorous approaches, with one example being PI-explanations. Unfortunately, PI-explanations also exhibit important drawbacks, the most visible of which is arguably their size. Recently, it has been observed that the (absolute) rigor of PI-explanations can be traded off for a smaller explanation size, by computing the so-called relevant sets. Given some positive {\delta}, a set S of features is {\delta}-relevant if, when the features in S are fixed, the probability of getting the target class exceeds {\delta}. However, even for very simple classifiers, the complexity of computing relevant sets of features is prohibitive, with the decision problem being NPPP-complete for circuit-based classifiers. In contrast with earlier negative results, this paper investigates practical approaches for computing relevant sets for a number of widely used classifiers that include Decision Trees (DTs), Naive Bayes Classifiers (NBCs), and several families of classifiers obtained from propositional languages. Moreover, the paper shows that, in practice, and for these families of classifiers, relevant sets are easy to compute. Furthermore, the experiments confirm that succinct sets of relevant features can be obtained for the families of classifiers considered., Comment: arXiv admin note: text overlap with arXiv:2207.04748, arXiv:2205.09569

Excitons and free charge carriers with large wave vectors form a nonradiative reservoir, which can strongly affect properties of bright excitons due to the exciton-exciton and exciton-carrier interactions. In the present work, the dynamics of quasiparticles in the reservoir at different areal densities is experimentally studied in a GaAs/AlGaAs quantum well using a time-resolved reflectance spectroscopy of nonradiative broadening of exciton resonances. The population of the reservoir is controlled either by the excitation power or by the temperature of the structure under study. The dynamics is quantitatively analyzed in the framework of the model developed earlier [Kurdyubov et al., Phys. Rev. B {\bf 104}, 035414 (2021)]. The model considers several dynamic processes, such as scattering of photoexcited excitons into the reservoir, dissociation of excitons into free charge carriers and the reverse process, carrier-induced exciton scattering into the light cone depleting the reservoir, thermally activated nonradiative losses of charge carriers. We have found that competition of these processes leads to highly nonlinear dynamics of reservoir excitons, although their density is far below the exciton Mott transition densities. Characteristic rates of the processes and their dependencies on the excitation power and the temperature are determined., Comment: Submitted to PRB

The rise of AI methods to make predictions and decisions has led to a pressing need for more explainable artificial intelligence (XAI) methods. One common approach for XAI is to produce a post-hoc explanation, explaining why a black box ML model made a certain prediction. Formal approaches to post-hoc explanations provide succinct reasons for why a prediction was made, as well as why not another prediction was made. But these approaches assume that features are independent and uniformly distributed. While this means that "why" explanations are correct, they may be longer than required. It also means the "why not" explanations may be suspect as the counterexamples they rely on may not be meaningful. In this paper, we show how one can apply background knowledge to give more succinct "why" formal explanations, that are presumably easier to interpret by humans, and give more accurate "why not" explanations. In addition, we show how to use existing rule induction techniques to efficiently extract background information from a dataset, and also how to report which background information was used to make an explanation, allowing a human to examine it if they doubt the correctness of the explanation.

Decision trees (DTs) epitomize the ideal of interpretability of machine learning (ML) models. The interpretability of decision trees motivates explainability approaches by so-called intrinsic interpretability, and it is at the core of recent proposals for applying interpretable ML models in high-risk applications. The belief in DT interpretability is justified by the fact that explanations for DT predictions are generally expected to be succinct. Indeed, in the case of DTs, explanations correspond to DT paths. Since decision trees are ideally shallow, and so paths contain far fewer features than the total number of features, explanations in DTs are expected to be succinct, and hence interpretable. This paper offers both theoretical and experimental arguments demonstrating that, as long as interpretability of decision trees equates with succinctness of explanations, then decision trees ought not be deemed interpretable. The paper introduces logically rigorous path explanations and path explanation redundancy, and proves that there exist functions for which decision trees must exhibit paths with arbitrarily large explanation redundancy. The paper also proves that only a very restricted class of functions can be represented with DTs that exhibit no explanation redundancy. In addition, the paper includes experimental results substantiating that path explanation redundancy is observed ubiquitously in decision trees, including those obtained using different tree learning algorithms, but also in a wide range of publicly available decision trees. The paper also proposes polynomial-time algorithms for eliminating path explanation redundancy, which in practice require negligible time to compute. Thus, these algorithms serve to indirectly attain irreducible, and so succinct, explanations for decision trees.

Decision trees (DTs) embody interpretable classifiers. DTs have been advocated for deployment in high-risk applications, but also for explaining other complex classifiers. Nevertheless, recent work has demonstrated that predictions in DTs ought to be explained with rigorous approaches. Although rigorous explanations can be computed in polynomial time for DTs, their size may be beyond the cognitive limits of human decision makers. This paper investigates the computation of {\delta}-relevant sets for DTs. {\delta}-relevant sets denote explanations that are succinct and provably precise. These sets represent generalizations of rigorous explanations, which are precise with probability one, and so they enable trading off explanation size for precision. The paper proposes two logic encodings for computing smallest {\delta}-relevant sets for DTs. The paper further devises a polynomial-time algorithm for computing {\delta}-relevant sets which are not guaranteed to be subset-minimal, but for which the experiments show to be most often subset-minimal in practice. The experimental results also demonstrate the practical efficiency of computing smallest {\delta}-relevant sets.

Thermal tides in the Martian atmosphere are analyzed using temperature profiles retrieved from nadir observations obtained by the TIRVIM Fourier-spectrometer, part of the Atmospheric Chemistry Suite (ACS) onboard the ExoMars Trace Gas Orbiter (TGO). The data is selected near the northern summer solstice at solar longitude (LS) 75{\deg}-105{\deg} of Martian Year (MY) 35. The observations have a full local time coverage, which enables analyses of daily temperature anomalies. The observed zonal mean temperature is lower by 4-6K at ~100Pa, but higher towards the summer pole, compared to the LMD Mars General Circulation Model (GCM). Wave mode decomposition shows dominant diurnal tide and important semi-diurnal tide and diurnal Kelvin wave, with maximal amplitudes of 5K, 3K, and 2.5K, respectively, from tens to hundreds of Pa. The results generally agree well with the LMD Mars GCM, but with noticeable earlier phases of diurnal (~1h) and semi-diurnal (~3h) tides., Comment: 7 pages, 4 figures, accepted by GRL

The Atmospheric Chemistry Suite (ACS) onboard the ExoMars Trace Gas Orbiter (TGO) monitors the Martian atmosphere through different spectral intervals in the infrared light. We present a retrieval algorithm tailored to the analysis of spectra acquired in nadir geometry by TIRVIM, the thermal infrared channel of ACS. Our algorithm simultaneously retrieves vertical profile of atmospheric temperature up to 50 km, surface temperature, and integrated optical depth of dust and water ice clouds. The specificity of the TIRVIM dataset lies in its capacity to resolve the diurnal cycle over a 54 sol period. However, it is uncertain to what extent can the desired atmospheric quantities be accurately estimated at different times of day. Here we first present an Observing System Simulation Experiment (OSSE). We produce synthetic observations at various latitudes, seasons and local times and run our retrieval algorithm on these synthetic data, to evaluate its robustness. Different sources of biases are documented, in particular regarding aerosol retrievals. Atmospheric temperature retrievals are found robust even when dust and/or water ice cloud opacities are not well estimated in our OSSE. We then apply our algorithm to TIRVIM observations in April-May, 2018 and perform a cross-validation of retrieved atmospheric temperature and dust integrated opacity by comparisons with thousands of co-located Mars Climate Sounder (MCS) retrievals. Most differences between TIRVIM and MCS atmospheric temperatures can be attributed to differences in vertical sensitivity. Daytime dust opacities agree well with each other, while biases are found in nighttime dust opacity retrieved from TIRVIM at this season., Comment: 51 pages, 29 figures. Accepted in Juanuary 2022 for publication in Journal of Geophysical Research (Planets)

One of the most widespread human behavioral biases is the present bias -- the tendency to overestimate current costs by a bias factor. Kleinberg and Oren (2014) introduced an elegant graph-theoretical model of inconsistent planning capturing the behavior of a present-biased agent accomplishing a set of actions. The essential measure of the system introduced by Kleinberg and Oren is the cost of irrationality -- the ratio of the total cost of the actions performed by the present-biased agent to the optimal cost. This measure is vital for a task designer to estimate the aftermaths of human behavior related to time-inconsistent planning, including procrastination and abandonment. As we prove in this paper, the cost of irrationality is highly susceptible to the agent's choices when faced with a few possible actions of equal estimated costs. To address this issue, we propose a modification of Kleinberg-Oren's model of inconsistent planning. In our model, when an agent selects from several options of minimum prescribed cost, he uses a randomized procedure. We explore the algorithmic complexity of computing and estimating the cost of irrationality in the new model.

Wave functions of heavy-hole excitons in GaAs/Al$_{0.3}$Ga$_{0.7}$As square quantum wells (QWs) of various widths are calculated by the direct numerical solution of a three-dimensional Schr\"odinger equation using a finite-difference scheme. These wave functions are then used to determine the exciton-exciton, exciton-electron and exciton-hole fermion exchange constants in a wide range of QW widths (5-150 nm). Additionally, the spin-dependent matrix elements of elastic exciton-exciton, exciton-electron and exciton-hole scattering are calculated. From these matrix elements, the collisional broadening of the exciton resonance is obtained within the Born approximation as a function of the areal density of excitons, electrons and holes respectively for QW widths of 5, 15, 30 and 50 nm. The obtained numerical results are compared with other theoretical works., Comment: 15 pages, 3 figures

Knowledge compilation (KC) languages find a growing number of practical uses, including in Constraint Programming (CP) and in Machine Learning (ML). In most applications, one natural question is how to explain the decisions made by models represented by a KC language. This paper shows that for many of the best known KC languages, well-known classes of explanations can be computed in polynomial time. These classes include deterministic decomposable negation normal form (d-DNNF), and so any KC language that is strictly less succinct than d-DNNF. Furthermore, the paper also investigates the conditions under which polynomial time computation of explanations can be extended to KC languages more succinct than d-DNNF.

The near-infrared (NIR) spectral range (from 780 to 2500 nm) of the multispectral remote sensing imagery provides vital information for the landcover classification, especially concerning the vegetation assessment. Despite the usefulness of NIR, common RGB is not always accompanied by it. Modern achievements in image processing via deep neural networks allow generating artificial spectral information, such as for the image colorization problem. In this research, we aim to investigate whether this approach can produce not only visually similar images but also an artificial spectral band that can improve the performance of computer vision algorithms for solving remote sensing tasks. We study the generative adversarial network (GAN) approach in the task of the NIR band generation using just RGB channels of high-resolution satellite imagery. We evaluate the impact of a generated channel on the model performance for solving the forest segmentation task. Our results show an increase in model accuracy when using generated NIR comparing to the baseline model that uses only RGB (0.947 and 0.914 F1-score accordingly). Conducted study shows the advantages of generating the extra band and its implementation in applied challenges reducing the required amount of labeled data.

Recent work has shown that not only decision trees (DTs) may not be interpretable but also proposed a polynomial-time algorithm for computing one PI-explanation of a DT. This paper shows that for a wide range of classifiers, globally referred to as decision graphs, and which include decision trees and binary decision diagrams, but also their multi-valued variants, there exist polynomial-time algorithms for computing one PI-explanation. In addition, the paper also proposes a polynomial-time algorithm for computing one contrastive explanation. These novel algorithms build on explanation graphs (XpG's). XpG's denote a graph representation that enables both theoretical and practically efficient computation of explanations for decision graphs. Furthermore, the paper proposes a practically efficient solution for the enumeration of explanations, and studies the complexity of deciding whether a given feature is included in some explanation. For the concrete case of decision trees, the paper shows that the set of all contrastive explanations can be enumerated in polynomial time. Finally, the experimental results validate the practical applicability of the algorithms proposed in the paper on a wide range of publicly available benchmarks.

Recent work proposed $\delta$-relevant inputs (or sets) as a probabilistic explanation for the predictions made by a classifier on a given input. $\delta$-relevant sets are significant because they serve to relate (model-agnostic) Anchors with (model-accurate) PI- explanations, among other explanation approaches. Unfortunately, the computation of smallest size $\delta$-relevant sets is complete for ${NP}^{PP}$, rendering their computation largely infeasible in practice. This paper investigates solutions for tackling the practical limitations of $\delta$-relevant sets. First, the paper alternatively considers the computation of subset-minimal sets. Second, the paper studies concrete families of classifiers, including decision trees among others. For these cases, the paper shows that the computation of subset-minimal $\delta$-relevant sets is in NP, and can be solved with a polynomial number of calls to an NP oracle. The experimental evaluation compares the proposed approach with heuristic explainers for the concrete case of the classifiers studied in the paper, and confirms the advantage of the proposed solution over the state of the art.

In many classification tasks there is a requirement of monotonicity. Concretely, if all else remains constant, increasing (resp. decreasing) the value of one or more features must not decrease (resp. increase) the value of the prediction. Despite comprehensive efforts on learning monotonic classifiers, dedicated approaches for explaining monotonic classifiers are scarce and classifier-specific. This paper describes novel algorithms for the computation of one formal explanation of a (black-box) monotonic classifier. These novel algorithms are polynomial in the run time complexity of the classifier and the number of features. Furthermore, the paper presents a practically efficient model-agnostic algorithm for enumerating formal explanations.

Today deep convolutional neural networks (CNNs) push the limits for most computer vision problems, define trends, and set state-of-the-art results. In remote sensing tasks such as object detection and semantic segmentation, CNNs reach the SotA performance. However, for precise performance, CNNs require much high-quality training data. Rare objects and the variability of environmental conditions strongly affect prediction stability and accuracy. To overcome these data restrictions, it is common to consider various approaches including data augmentation techniques. This study focuses on the development and testing of object-based augmentation. The practical usefulness of the developed augmentation technique is shown in the remote sensing domain, being one of the most demanded ineffective augmentation techniques. We propose a novel pipeline for georeferenced image augmentation that enables a significant increase in the number of training samples. The presented pipeline is called object-based augmentation (OBA) and exploits objects' segmentation masks to produce new realistic training scenes using target objects and various label-free backgrounds. We test the approach on the buildings segmentation dataset with six different CNN architectures and show that the proposed method benefits for all the tested models. We also show that further augmentation strategy optimization can improve the results. The proposed method leads to the meaningful improvement of U-Net model predictions from 0.78 to 0.83 F1-score.

Dynamics of nonradiative excitons with large in-plane wave vectors forming a so-called reservoir is experimentally studied in a high-quality semiconductor structure containing a 14-nm shallow GaAs/Al$_{0.03}$Ga$_{0.97}$As quantum well by means of the non-degenerate pump-probe spectroscopy. The exciton dynamics is visualized via the dynamic broadening of the heavy-hole and light-hole exciton resonances caused by the exciton-exciton scattering. Under the non-resonant excitation free carriers are optically generated. In this regime the exciton dynamics is strongly affected by the exciton-carrier scattering. In particular, if the carriers of one sign are prevailing, they efficiently deplete the reservoir of the nonradiative excitons inducing their scattering into the light cone. A simple model of the exciton dynamics is developed, which considers the energy relaxation of photocreated electrons and holes, their coupling into excitons, and exciton scattering into the light cone. The model well reproduces the exciton dynamics observed experimentally both at the resonant and nonresonant excitation. Moreover, it correctly describes the profiles of the photoluminescence pulses studied experimentally. The efficient exciton-electron interaction is further experimentally verified by the control of the exciton density in the reservoir when an additional excitation creates electrons depleting the reservoir., Comment: 16 pages, 10 figures

In this work, a system for creating a relightable 3D portrait of a human head is presented. Our neural pipeline operates on a sequence of frames captured by a smartphone camera with the flash blinking (flash-no flash sequence). A coarse point cloud reconstructed via structure-from-motion software and multi-view denoising is then used as a geometric proxy. Afterwards, a deep rendering network is trained to regress dense albedo, normals, and environmental lighting maps for arbitrary new viewpoints. Effectively, the proxy geometry and the rendering network constitute a relightable 3D portrait model, that can be synthesized from an arbitrary viewpoint and under arbitrary lighting, e.g. directional light, point light, or an environment map. The model is fitted to the sequence of frames with human face-specific priors that enforce the plausibility of albedo-lighting decomposition and operates at the interactive frame rate. We evaluate the performance of the method under varying lighting conditions and at the extrapolated viewpoints and compare with existing relighting methods.

Decision trees (DTs) epitomize what have become to be known as interpretable machine learning (ML) models. This is informally motivated by paths in DTs being often much smaller than the total number of features. This paper shows that in some settings DTs can hardly be deemed interpretable, with paths in a DT being arbitrarily larger than a PI-explanation, i.e. a subset-minimal set of feature values that entails the prediction. As a result, the paper proposes a novel model for computing PI-explanations of DTs, which enables computing one PI-explanation in polynomial time. Moreover, it is shown that enumeration of PI-explanations can be reduced to the enumeration of minimal hitting sets. Experimental results were obtained on a wide range of publicly available datasets with well-known DT-learning tools, and confirm that in most cases DTs have paths that are proper supersets of PI-explanations.

Decision lists are one of the most easily explainable machine learning models. Given the renewed emphasis on explainable machine learning decisions, this machine learning model is increasingly attractive, combining small size and clear explainability. In this paper, we show for the first time how to construct optimal "perfect" decision lists which are perfectly accurate on the training data, and minimal in size, making use of modern SAT solving technology. We also give a new method for determining optimal sparse decision lists, which trade off size and accuracy. We contrast the size and test accuracy of optimal decisions lists versus optimal decision sets, as well as other state-of-the-art methods for determining optimal decision lists. We also examine the size of average explanations generated by decision sets and decision lists.