Automated Decision Procedures Research Articles

Automated Decisions and Article No. 22 GDPR of the European Union: An Analysis of the Right to an ‘Explanation’

The aim of this proposal is to analyze use of algorithms in automated decision procedures concerning subjective positions having legal relevance. In this regard, first paragraph of Article No. 22 GDPR, entitled Automated decision-making relating to natural persons, including profiling establishes that: “The data subject has right not to be subjected to a decision based solely on automated processing, including profiling, which produces legal effects that concern him or that significantly affects his person”. This norm gives rise to important questions concerning very nature of decision formulation through algorithms. In hypothesis, juridical nature or juridical effect of concept of “decision” could be restricted to circumstances such as financial solvency, predictability of motor vehicle accidents. In these cases, Right to refers to algorithmic decision obtained. For purposes of this abstract, most interesting element, which could give rise to a comparative judicial dispute, given validity of GDPR in all countries of European Union, concerns provision of par. 2 of Article No. 22, letters a) and c) which refers to implementation of to protect the rights, freedoms and legitimate interests of interested party”. What is counterpart used for implementing such measures? It would seem unquestionable that only element that can satisfy this appropriateness is human intervention, i.e. someone who has necessary authority, ability and competence to modify or revise decision disputed by user. Another authoritative source claims that such appropriate measures may also consist of automated systems that control algorithms, i.e. periodic reviews, introduction of procedures that verify accuracy of decision-making process or correct errors, discrimination issues, or inaccuracies from an outdated database in order to avoid self-learning algorithm based on wrong data and processes. Nevertheless, there are those who express perplexity about effectiveness of right to explanation, especially in relation to difficulty in interpreting legislative passages that envisage it within GDPR itself, together with doubts about mandatory nature of this explanatory right. This paper focuses on meaning of right to explanation of decision pursuant to Article No. 22 GDPR in order to allow interested parties to understand reasons for decision and possibly to dispute it.

HoTTSQL: proving query rewrites with univalent SQL semantics

Every database system contains a query optimizer that performs query rewrites. Unfortunately, developing query optimizers remains a highly challenging task. Part of the challenges comes from the intricacies and rich features of query languages, which makes reasoning about rewrite rules difficult. In this paper, we propose a machine-checkable denotational semantics for SQL, the de facto language for relational database, for rigorously validating rewrite rules. Unlike previously proposed semantics that are either non-mechanized or only cover a small amount of SQL language features, our semantics covers all major features of SQL, including bags, correlated subqueries, aggregation, and indexes. Our mechanized semantics, called HoTT SQL, is based on K-Relations and homotopy type theory, where we denote relations as mathematical functions from tuples to univalent types. We have implemented HoTTSQL in Coq, which takes only fewer than 300 lines of code and have proved a wide range of SQL rewrite rules, including those from database research literature (e.g., magic set rewrites) and real-world query optimizers (e.g., subquery elimination). Several of these rewrite rules have never been previously proven correct. In addition, while query equivalence is generally undecidable, we have implemented an automated decision procedure using HoTTSQL for conjunctive queries: a well studied decidable fragment of SQL that encompasses many real-world queries.

Sequence Diagram Aided Privacy Policy Specification

A fundamental problem in the specification of regulatory privacy policies such as the Health Insurance Portability and Accountability Act (HIPAA) in a computer system is to state the policies precisely, consistent with their high-level intuition. In this paper, we propose UML sequence diagrams as a practical means to graphically express privacy policies. A graphical representation allows decision-makers such as application domain experts and security architects to easily verify and confirm the expected behavior. Once intuitively confirmed, our work in this article introduces an algorithmic approach to formalizing the semantics of sequence diagrams in terms of linear temporal logic (LTL) templates. In all the templates, different semantic aspects are expressed as separate, yet simple LTL formulas that can be composed to define the complex semantics of sequence diagrams. The formalization enables us to leverage the analytical powers of automated decision procedures for LTL formulas to determine if a collection of sequence diagrams is consistent, independent, etc. and also to verify if a system design conforms to the privacy policies. We evaluate our approach by modeling and analyzing a substantial subset of HIPAA rules using sequence diagrams.

Data-Based Multiobjective Plant-Wide Performance Optimization of Industrial Processes Under Dynamic Environments

This paper provides a method for automatically selecting optimal operational indices for unit processes in an industrial plant using measured data and without knowing dynamical models of the unit process. A dynamic multiobjective optimization problem is defined to find operational indices that lead to plant-wide production indices close to their target values. A case-based reasoning (CBR) technique is also employed, which uses the stored experience of a human expert to determine appropriate operational indices for given target production indices. The solutions of the optimization problem and CBR technique are combined to form baseline operational indices. The dynamic models of the production indices, however, are time varying and affected by disturbances and online corrections of these baseline operational indices are required. To this end, reinforcement learning (RL) is used to provide a data-driven optimization technique to compensate for disturbances and model approximation errors and variations. The data-driven RL approach is used in two different time scales. The samples of the predicted production indices are used at a fast sampling rate, i.e., at each sample time, and the samples of actual production indices are used at a slower sampling rate, i.e., after each operational run, to correct the baseline operational indices. The effectiveness of this automated decision procedure has been demonstrated by successful implementation of the proposed approach on a large mineral processing plant in Gansu Province, China.

A Refinement-Based Compositional Reasoning Framework for Pipelined Machine Verification

We present a refinement-based compositional framework for showing that pipelined machines satisfy the same safety and liveness properties as their non-pipelined specifications. Our framework consists of a set of convenient, easily applicable, and complete compositional proof rules. We show how to apply our compositional framework in the context of microprocessor verification to verify both abstract, term-level models and executable, bit-level models. Our framework enables us to verify machine models that are significantly more complex than the kinds of models that can be verified using current state-of-the-art automated decision procedures. For example, using our framework, we can verify a 32-bit, 10-stage, executable pipelined machine model. In addition, our compositional framework offers drastic improvements in the context of design debugging over monolithic approaches, in part because bugs are isolated to particular steps in the compositional proof and because the counter examples generated are much smaller.

Automated decision procedure for earthquake early warning

Early warning systems represent an effective tool for disseminating timely information about potentially catastrophic hazards to the public, emergency managers, and operators and users of transportation systems, industrial facilities, public buildings including schools and government offices, and private residences. This information gives the possibility of taking action to initiate mitigation or security measures before a catastrophic event occurs. Earthquake early warning technologies represent a promising application, especially in areas where faults are located in the proximity of an urbanized area, such as in Los Angeles and San Francisco, California. For earthquake early warning systems, automated procedures for taking action are necessary due to the short lead time available, including an automated decision procedure to decide when to activate mitigation or security measures based on the expected consequences of taking, or not taking, such actions. In this work, a decision procedure is presented that accounts for the uncertainty in the predicted ground shaking intensity at a site of a facility by deciding whether or not to take a mitigation action based on the probabilities of false and missed alarms. It uses a cost-benefit analysis of the consequences of making a wrong decision. It is shown how the method may be extended to other predictors, such as expected losses, that may be more effective parameters for rational decision making from the perspective of facility stakeholders.

Rewrite rule systems for modal propositional logic

This paper explains new results relating modal propositional logic and rewrite rule systems. More precisely, we give complete term rewriting systems for the modal propositional systems known as K, Q, T, and S5. These systems are presented as extensions of Hsiang's system for classical propositional calculus. We have checked local confluence with the rewrite rule system K.B. (cf. the Knuth-Bendix algorithm) developed by the Formel project at INRIA. We prove that these systems are noetherian, and then infer their confluence from Newman's lemma. Therefore each term rewriting system provides a new automated decision procedure and defines a canonical form for the corresponding logic. We also show how to characterize the canonical forms thus obtained.

Auditory‐nerve discharge characteristics that distinguish between basilar papilla neurons and amphibian papilla neurons in the leopard frog, Rana pipiens

The frog's inner ear has two sensory organs for auditory reception. One organ, the amphibian papilla (ap), has many of the complexities of mammalian ears. The other organ, the basilar papilla (bp), is much simpler and offers special opportunities for comparison to the mammalian inner hair cell configuration. These anatomical differences make it important to identify the site of the receptor cell when recording from individual afferent neurons in the main trunk of the auditory nerve. In routine experiments, this can be a difficult problem, since the characteristic frequency (cf) of the highest‐frequency (ap) neurons overlaps the cf of bp neurons in many species. To approach this problem, a statistical data base was created using responses to tone bursts and clicks, tuning curve characteristics, and spontaneous rates. A standard statistical method, discriminant analysis, produced reliable classification of neurons without using cf estimates and offers the possibility of a fully automated decision procedure. [Work supported by NIH grants NS21592 and RR01379.]