Search

Machine learning black boxes, exemplified by deep neural networks, often exhibit challenges in interpretability due to their reliance on complicated relationships involving numerous internal parameters and input features. This lack of transparency from a human perspective renders their predictions untrustworthy, particularly in critical applications. In this paper, we address this issue by introducing the design and implementation of CReEPy, an algorithm for symbolic knowledge extraction based on explainable clustering. Specifically, CReEPy leverages the underlying clustering performed by the ExACT or CREAM algorithms to generate human-interpretable Prolog rules that mimic the behaviour of opaque models. Additionally, we introduce CRASH, an algorithm for the automated tuning of hyper-parameters required by CReEPy. We present experiments evaluating both the human readability and predictive performance of the proposed knowledge-extraction algorithm, employing existing state-of-the-art techniques as benchmarks for comparison in real-world applications.

Cubic gold-platinum test-masses (TMs) play the role of free-falling geodesic reference and interferometer end mirrors of the future Laser Interferometer Space Antenna (LISA) observatory for low-frequency gravitational wave detection in space. A similar arrangement has been proposed for the Chinese missions Taiji and TianQin and for the LISA follow-on missions such as ALIA and BBO. The TMs are charged by high-energy particles and photons. The deposited charge couples with stray electric fields surrounding the TMs thus inducing spurious forces that limit the sensitivity of the mission mainly at low frequencies. The TM charging was measured in space in 2016-2017 with LISA Pathfinder (LPF), meant for the testing of the LISA instrumentation. Unfortunately, during the time LPF remained in orbit, no solar energetic particle events or major astrophysical phenomena were observed. We aim at estimating the LISA TM charging attributable to long, short gamma-ray bursts (GRBs) and magnetar flares in comparison to that of charged particles of galactic and solar origin. The contribution of these major astrophysical phenomena to the LISA TM charging is discussed here for the first time. The results found here can be extended to LISA-like missions. The response of the radiation monitors hosted on the three LISA S/C is also evaluated. We show that long, intense extragalactic GRBs and galactic magnetar flares at kpc distances can be detected and monitored through a sudden change from positive to negative charging of the TMs and an increase of the TM charging noise. This is a unique signature since both galactic and solar particles charge positively the LISA TMs. This peculiar behavior of the TM charging would allow monitoring the whole dynamics of GRBs. The suggestion reported in the literature, about the detection of long GRBs and gravitational waves from the same sources, in principle, may apply to LISA and other LISA-like missions since the increase of the TM charging noise during these extreme transient phenomena is estimated to remain below the mission sensitivity while particle detectors are expected to saturate.

Hydrogenated amorphous silicon (a-Si:H) is a material having an intrinsically high radiation hardness that can be deposited on flexible substrates such as polyimide (PI). For these properties, a-Si:H can be used for the production of flexible sensors. a-Si:H sensors can be successfully utilized in dosimetry, beam monitoring for particle physics (X-ray, electron, gamma ray, and proton detection) and radiotherapy, radiation flux measurement for space applications (study of solar energetic particles and stellar events), and neutron flux measurements. In this article, we have studied the dosimetric X-ray response of n-i-p diodes deposited on PI. We measured the linearity of the photocurrent response to X-rays versus dose rate from which we have extracted the dosimetric X-ray sensitivity at various bias voltages. In particular, low bias voltage operation has been studied to assess the high energy efficiency of these kinds of sensors. A measurement of stability of X-ray response versus time has been shown. The effect of detectors annealing has been studied. Operation under bending at various bending radii is also shown.

As opaque decision systems are being increasingly adopted in almost any application field, issues about their lack of transparency and human readability are a concrete concern for end-users. Amongst existing proposals to associate human-interpretable knowledge with accurate predictions provided by opaque models, there are rule extraction techniques, capable of extracting symbolic knowledge out of opaque models. The quantitative assessment of the extracted knowledge’s quality is still an open issue. For this reason, we provide here a first approach to measure the knowledge quality, encompassing several indicators and providing a compact score reflecting readability, completeness and predictive performance associated with a symbolic knowledge representation. We also discuss the main criticalities behind our proposal, related to the readability assessment and evaluation, to push future research efforts towards a more robust score formulation.

The XAI community is currently studying and developing symbolic knowledge-extraction (SKE) algorithms as a means to produce human-intelligible explanations for black-box machine learning predictors, so as to achieve believability in human-machine interaction. However, many extraction procedures exist in the literature, and choosing the most adequate one is increasingly cumbersome, as novel methods keep on emerging. Challenges arise from the fact that SKE algorithms are commonly defined based on theoretical assumptions that typically hinder practical applicability. This paper focuses on hypercube-basedSKE methods, a quite general class of extraction techniques mostly devoted to regression-specific tasks. We first show that hypercube-based methods are flexible enough to support classification problems as well, then we propose a general model for them, and discuss how they support SKE on datasets, predictors, or learning tasks of any sort. Empirical examples are reported as well –based upon the PSyKE framework –, showing the applicability of hypercube-based methods to actual classification tasks.

A common practice in modern explainable AI is to post-hocexplain black-box machine learning (ML) predictors – such as neural networks – by extracting symbolicknowledge out of them, in the form of either rule lists or decision trees. By acting as a surrogate model, the extracted knowledge aims at revealing the inner working of the black box, thus enabling its inspection, representation, and explanation. Various knowledge-extraction algorithms have been presented in the literature so far. Unfortunately, running implementations of most of them are currently either proofs of concept or unavailable. In any case, a unified, coherent software framework supporting them all – as well as their interchange, comparison, and exploitation in arbitrary ML workflows – is currently missing. Accordingly, in this paper we discuss the design of PSyKE, a platform providing general-purpose support to symbolic knowledge extraction from different sorts of black-box predictors via many extraction algorithms. Notably, PSyKE targets symbolic knowledge in logic form, allowing the extraction of first-order logic clauses. The extracted knowledge is thus both machine- and human-interpretable, and can be used as a starting point for further symbolic processing—e.g. automated reasoning.