Machine Learning

One of the primary, if not most critical, difficulties in the design and implementation of autonomous systems is the black-boxed nature of the decision-making structures and logical pathways of autonomous systems. For this reason, the values of stakeholders become of particular significance given the risks posed by opaque structures of intelligent agents (IAs). This paper proposes the Value Sensitive Design (VSD) approach as a principled framework for incorporating these values in design. The example of autonomous vehicles is used as a (...) case study for how VSD offers a systematic way for engineering teams to formally incorporate existing technical solutions towards ethical design, while simultaneously remaining pliable to emerging issues and needs. It is concluded that the VSD methodology offers at least a strong enough foundation from which designers can begin to anticipate design needs and formulate salient design flows that can adapted to changing ethical landscapes. (shrink)

At its strongest, Hume's problem of induction denies the existence of any well justified assumptionless inductive inference rule. At the weakest, it challenges our ability to articulate and apply good inductive inference rules. This paper examines an analysis that is closer to the latter camp. It reviews one answer to this problem drawn from the VC theorem in statistical learning theory and argues for its inadequacy. In particular, I show that it cannot be computed, in general, whether we are in (...) a situation where the VC theorem can be applied for the purpose we want it to. (shrink)

A review of Gary Smith and Jay Cordes: The Phantom Pattern Problem: The Mirage of Big Data. New York: Oxford University Press, 2020.

As our epistemic ambitions grow, the common and scientific endeavours are becoming increasingly dependent on Machine Learning (ML). The field rests on a single experimental paradigm, which consists of splitting the available data into a training and testing set and using the latter to measure how well the trained ML model generalises to unseen samples. If the model reaches acceptable accuracy, an a posteriori contract comes into effect between humans and the model, supposedly allowing its deployment to target environments. Yet (...) the latter part of the contract depends on human inductive predictions or generalisations, which infer a uniformity between the trained ML model and the targets. The paper asks how we justify the contract between human and machine learning. It is argued that the justification becomes a pressing issue when we use ML to reach ‘elsewheres’ in space and time or deploy ML models in non-benign environments. The paper argues that the only viable version of the contract can be based on optimality (instead of on reliability which cannot be justified without circularity) and aligns this position with Schurz’s optimality justification. It is shown that when dealing with inaccessible/unstable ground-truths (‘elsewheres’ and non-benign targets), the optimality justification undergoes a slight change, which should reflect critically on our epistemic ambitions. Therefore, the study of ML robustness should involve not only heuristics that lead to acceptable accuracies on testing sets. The justification of human inductive predictions or generalisations about the uniformity between ML models and targets should be included as well. Without it, the assumptions about inductive risk minimisation in ML are not addressed in full. (shrink)

After generalizing the Archimedean property of real numbers in such a way as to make it adaptable to non-numeric structures, we demonstrate that the real numbers cannot be used to accurately measure non-Archimedean structures. We argue that, since an agent with Artificial General Intelligence (AGI) should have no problem engaging in tasks that inherently involve non-Archimedean rewards, and since traditional reinforcement learning rewards are real numbers, therefore traditional reinforcement learning probably will not lead to AGI. We indicate two possible ways (...) traditional reinforcement learning could be altered to remove this roadblock. (shrink)

The environmental costs and energy constraints have become emerging issues for the future development of Machine Learning (ML) and Artificial Intelligence (AI). So far, the discussion on environmental impacts of ML/AI lacks a perspective reaching beyond quantitative measurements of the energy-related research costs. Building on the foundations laid down by Schwartz et al., 2019 in the GreenAI initiative, our argument considers two interlinked phenomena, the gratuitous generalisation capability and the future where ML/AI performs the majority of quantifiable inductive inferences. The (...) gratuitous generalisation capability refers to a discrepancy between the cognitive demands of a task to be accomplished and the performance (accuracy) of a used ML/AI model. If the latter exceeds the former because the model was optimised to achieve the best possible accuracy, it becomes inefficient and its operation harmful to the environment. The future dominated by the non-anthropic induction describes a use of ML/AI so all-pervasive that most of the inductive inferences become furnished by ML/AI generalisations. The paper argues that the present debate deserves an expansion connecting the environmental costs of research and ineffective ML/AI uses (the issue of gratuitous generalisation capability) with the (near) future marked by the all-pervasive Human-Artificial Intelligence Nexus. (shrink)

The paper proposes a synthesis between human scientists and artificial representation learning models as a way of augmenting epistemic warrants of realist theories against various anti-realist attempts. Towards this end, the paper fleshes out unconceived alternatives not as a critique of scientific realism but rather a reinforcement, as it rejects the retrospective interpretations of scientific progress, which brought about the problem of alternatives in the first place. By utilising adversarial machine learning, the synthesis explores possibility spaces of available evidence for (...) unconceived alternatives providing modal knowledge of what is possible therein. As a result, the epistemic warrant of synthesised realist theories should emerge bolstered as the underdetermination by available evidence gets reduced. While shifting the realist commitment away from theoretical artefacts towards modalities of the possibility spaces, the synthesis comes out as a kind of perspectival modelling. (shrink)

几年前,我通常可以从书名中分辨出什么,或者至少从章节标题中看出,会犯什么样的哲学错误,以及错误的频率。就名义上的科学著作而言,这些可能在很大程度上局限于某些章节,这些章节具有哲学意义或试图得出关于该作 品的意义或长期意义的一般性结论。然而,通常情况下,事实的科学问题慷慨地与哲学的胡言乱语,这些事实意味着什么。维特根斯坦在大约80年前描述的科学问题与各种语言游戏所描述的明确区别很少被考虑,因此人们交替 地被科学所震惊,并因它的不连贯而感到沮丧。分析。因此,这是与这个卷。 如果一个人要创造一个或多或少像我们一样的头脑,一个人需要有一个理性的逻辑结构,并理解两种思想体系(双过程理论)。如果一个人要对此进行哲学思考,就需要理解科学事实问题与语言如何在问题语境中工作,以及如何 避免还原主义和科学主义的陷阱的哲学问题之间的区别,但Kurzweil,如最学生的行为,基本上都是无知的。他被模型、理论和概念所陶醉,以及解释的冲动,而维特根斯坦向我们表明,我们只需要描述,理论、概念等 只是使用语言(语言游戏)的方式,只有它们有明确的价值测试(清晰的真理制造者,或约翰西尔(AI最著名的批评家)喜欢说,明确的满意条件(COS))。我试图在我最近的著作中对此作一个开端。 那些希望从现代两个系统的观点来看为人类行为建立一个全面的最新框架的人,可以查阅我的书《路德维希的哲学、心理学、Mind 和语言的逻辑结构》维特根斯坦和约翰·西尔的《第二部》(2019年)。那些对我更多的作品感兴趣的人可能会看到《会说话的猴子——一个末日星球上的哲学、心理学、科学、宗教和政治——文章和评论2006-201 9年第3次(2019年)和自杀乌托邦幻想21篇世纪4日 (2019) .

Despite the seemingly neutral vantage of using nature for widely-distributed computational purposes, neither post-biological nor post-humanist teleology simply concludes with the real "end of nature" as entailed in the loss of the specific ontological status embedded in the identifier "natural." As evinced by the ecological crises of the Anthropocene—of which the 2019 Brazil Amazon rainforest fires are only the most recent—our epoch has transfixed the “natural order" and imposed entropic artificial integration, producing living species that become “anoetic,” made to serve (...) as automated exosomatic residues, or digital flecks. I further develop Gilles Deleuze’s description of control societies to upturn Foucauldian biopower, replacing its spacio-temporal bounds with the exographic excesses in psycho-power; culling and further detailing Bernard Stiegler’s framework of transindividuation and hyper-control, I examine how becoming-subject is predictively facilitated within cognitive capitalism and what Alexander Galloway terms “deep digitality.” Despite the loss of material vestiges qua virtualization—which I seek to trace in an historical review of industrialization to postindustrialization—the drive-based and reticulated "internet of things" facilitates a closed loop from within the brain to the outside environment, such that the aperture of thought is mediated and compressed. The human brain, understood through its material constitution, is susceptible to total datafication’s laminated process of “becoming-mnemotechnical,” and, as neuroplasticity is now a valid description for deep-learning and neural nets, we are privy to the rebirth of the once-discounted metaphor of the “cybernetic brain.” Probing algorithmic governmentality while posing noetic dreaming as both technical and pharmacological, I seek to analyze how spirit is blithely confounded with machine-thinking’s gelatinous cognition, as prosthetic organ-adaptation becomes probabilistically molded, networked, and agentially inflected (rather than simply externalized). (shrink)

Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which they (...) are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy. (shrink)

Artificial intelligence (AI) research enjoyed an initial period of enthusiasm in the 1970s and 80s. But this enthusiasm was tempered by a long interlude of frustration when genuinely useful AI applications failed to be forthcoming. Today, we are experiencing once again a period of enthusiasm, fired above all by the successes of the technology of deep neural networks or deep machine learning. In this paper we draw attention to what we take to be serious problems underlying current views of artificial (...) intelligence encouraged by these successes, especially in the domain of language processing. We then show an alternative approach to language-centric AI, in which we identify a role for philosophy. (shrink)

An important problem with machine learning is that when label number n>2, it is very difficult to construct and optimize a group of learning functions, and we wish that optimized learning functions are still useful when prior distribution P(x) (where x is an instance) is changed. To resolve this problem, the semantic information G theory, Logical Bayesian Inference (LBI), and a group of Channel Matching (CM) algorithms together form a systematic solution. MultilabelMultilabel A semantic channel in the G theory consists (...) of a group of truth functions or membership functions. In comparison with likelihood functions, Bayesian posteriors, and Logistic functions used by popular methods, membership functions can be more conveniently used as learning functions without the above problem. In Logical Bayesian Inference (LBI), every label’s learning is independent. For Multilabel learning, we can directly obtain a group of optimized membership functions from a big enough sample with labels, without preparing different samples for different labels. A group of Channel Matching (CM) algorithms are developed for machine learning. For the Maximum Mutual Information (MMI) classification of three classes with Gaussian distributions on a two-dimensional feature space, 2-3 iterations can make mutual information between three classes and three labels surpass 99% of the MMI for most initial partitions. For mixture models, the Expectation-Maxmization (EM) algorithm is improved and becomes the CM-EM algorithm, which can outperform the EM algorithm when mixture ratios are imbalanced, or local convergence exists. The CM iteration algorithm needs to combine neural networks for MMI classifications on high-dimensional feature spaces. LBI needs further studies for the unification of statistics and logic. (shrink)

By drawing on the philosophy of Bernard Stiegler, the phenomena of mechanical (a.k.a. artificial, digital, or electronic) intelligence is explored in terms of its real significance as an ever-repeating threat of the reemergence of stupidity (as cowardice), which can be transformed into knowledge (pharmacological analysis of poisons and remedies) by practices of care, through the outlook of what researchers describe equivocally as “artificial stupidity”, which has been identified as a new direction in the future of computer science and machine problem (...) solving as well as a new difficulty to be overcome. I weave together of web of “artificial stupidity”, which denotes the mechanic (1), the human (2), or the global (3). With regards to machine intelligence, artificial stupidity refers to: 1a) Weak A.I. or a rhetorical inversion of designating contemporary practices of narrow task-based procedures by algorithms in opposition to “True A.I.”; 1b) the restriction or employment of constraints that weaken the effectiveness of A.I., which is to say a “dumbing-down” of A.I. by intentionally introducing mistakes by programmers for safety concerns and human interaction purposes; 1c) the failure of machines to perform designated tasks; 1d) a lack of a noetic capacity, which is a lack of moral and ethical discretion; 1e) a lack of causal reasoning (true intelligence) as opposed to statistical associative “curve fitting”; or 2) the phenomenon of increasing human “stupidity” or drive-based behaviors, which is considered as the degradation of human intelligence and/or “intelligent human behavior” through technics; and finally, 3) the global phenomenon of increasing entropy due to a black-box economy of closed systems and/or industry consolidation. (shrink)

In this paper, a predictive artificial neural network (ANN) model was developed and validated for the purpose of prediction whether a watermelon is good or bad, the model was developed using JUSTNN software environment. Prediction is done based on some watermelon attributes that are chosen to be input data to the ANN. Attributes like color, density, sugar rate, and some others. The model went through multiple learning-validation cycles until the error is zero, so the model is 100% percent accurate for (...) the purpose of prediction whether good or bad the watermelon is. (shrink)

In this paper, an artificial neural network (ANN) model was developed and validated to predict whether a couple is going to get divorced or not. Prediction is done based on some questions that the couple answered, answers of those questions were used as the input to the ANN. The model went through multiple learning-validation cycles until it got 100% accuracy.

In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI. Intuitively, the purpose of providing an explanation of a model or a decision is to make it understandable to its stakeholders. But without a previous grasp of what it means to say that an agent understands a model or a decision, the explanatory strategies will (...) lack a well-defined goal. Aside from providing a clearer objective for XAI, focusing on understanding also allows us to relax the factivity condition on explanation, which is impossible to fulfill in many machine learning models, and to focus instead on the pragmatic conditions that determine the best fit between a model and the methods and devices deployed to understand it. After an examination of the different types of understanding discussed in the philosophical and psychological literature, I conclude that interpretative or approximation models not only provide the best way to achieve the objectual understanding of a machine learning model, but are also a necessary condition to achieve post hoc interpretability. This conclusion is partly based on the shortcomings of the purely functionalist approach to post hoc interpretability that seems to be predominant in most recent literature. (shrink)

When philosophers discuss the possibility of machines making scientific discoveries, they typically focus on discoveries in physics, biology, chemistry and mathematics. Observing the rapid increase of computer-use in science, however, it becomes natural to ask whether there are any scientific domains out of reach for machine discovery. For example, could machines also make discoveries in qualitative social science? Is there something about humans that makes us uniquely suited to studying humans? Is there something about machines that would bar them from (...) such activity? A close look at the methodology of interpretive social science reveals several abilities necessary to make a social scientific discovery, and one capacity necessary to possess any of them is imagination. For machines to make discoveries in social science, therefore, they must possess imagination algorithms. (shrink)

This paper shows the smart walking stick based on ultrasonic sensors and Arduino for outwardly debilitated individuals. There are roughly 37 million individuals over the globe who are visually impaired as indicated by the World Health Organization. Individuals with visual inabilities are regularly subjected to outer help which can be given by people, trained dogs, or electronic gadgets as supportive networks for basic assistance. Thus, this played as the motivation to develop a smart cane white stick to survive these restrictions (...) which includes ultrasonic sensors at particular positions to the stick that gave data about nature to the client by initiating the signal sound and vibrations. We proposed minimal effort and light weight framework structured with micro controller that processes signal and alerts the visually impaired person over any obstacle, water or dark areas through beeping sounds or vibrations. The system comprises of obstacle and moisture detection sensors for process of receiving, processing and sending signals to the alarm system which finally alerts the user for prompt action. The system was designed, programmed using Java language and tried for exactness by the visually impaired. Our gadget can recognize obstacle inside the separation of around 2m from the client. (shrink)

Lake et al. propose three criteria which, they argue, will bring artificial intelligence (AI) systems closer to human cognitive abilities. In this paper, we explore the application of these criteria to a particular domain of human cognition: our capacity for moral reasoning. In doing so, we explore a set of considerations relevant to the development of AI moral decision-making. Our main focus is on the relation between dual-process accounts of moral reasoning and model-free/model-based forms of machine learning. We also discuss (...) how work in embodied and situated cognition could provide a valu- able perspective on future research. (shrink)

While representation learning techniques have shown great promise in application to a number of different NLP tasks, they have had little impact on the problem of ontology matching. Unlike past work that has focused on feature engineering, we present a novel representation learning approach that is tailored to the ontology matching task. Our approach is based on embedding ontological terms in a high-dimensional Euclidean space. This embedding is derived on the basis of a novel phrase retrofitting strategy through which semantic (...) similarity information becomes inscribed onto fields of pre-trained word vectors. The resulting framework also incorporates a novel outlier detection mechanism based on a denoising autoencoder that is shown to improve performance. An ontology matching system derived using the proposed framework achieved an F-score of 94% on an alignment scenario involving the Adult Mouse Anatomical Dictionary and the Foundational Model of Anatomy ontology (FMA) as targets. This compares favorably with the best performing systems on the Ontology Alignment Evaluation Initiative anatomy challenge. We performed additional experiments on aligning FMA to NCI Thesaurus and to SNOMED CT based on a reference alignment extracted from the UMLS Metathesaurus. Our system obtained overall F-scores of 93.2% and 89.2% for these experiments, thus achieving state-of-the-art results. (shrink)

ITS (Intelligent Tutoring System) is a computer software that supplies direct and adaptive training or response to students without, or with little human teacher interfering. The main target of ITS is smoothing the learning-teaching process using the ultimate technology in computer science. The proposed system will be implemented using the “ITSB” Authoring tool. The book "Introduction To Computer Science" is taught in Al-Azhar University in Gaza as a compulsory subject for students who study at humanities faculties. In this thesis, the (...) researcher demonstrates an intelligent tutoring system for teaching the above mentioned subject. The system was assessed by a group of teachers and students and the results were promising . (shrink)

Educational Data Mining (EDM) is a new paradigm aiming to mine and extract knowledge necessary to optimize the effectiveness of teaching process. With normal educational system work it’s often unlikely to accomplish fine system optimizing due to large amount of data being collected and tangled throughout the system. EDM resolves this problem by its capability to mine and explore these raw data and as a consequence of extracting knowledge. This paper describes several experiments on real educational data wherein the effectiveness (...) of Data Mining is explained in migration the educational data into knowledge. The experiments goal at first to identify important factors of teacher behaviors influencing student satisfaction. In addition to presenting experiences gained through the experiments, the paper aims to provide practical guidance of Data Mining solutions in a real application. (shrink)

The orbital space environment is home to natural and artificial satellites, debris, and space weather phenomena. As the population of orbital objects grows so do the potential hazards to astronauts, space infrastructure and spaceflight capability. Orbital debris, in particular, is a universal concern. This and other hazards can be minimized by improving global space situational awareness (SSA). By sharing more data and increasing observational coverage of the space environment we stand to achieve that goal, thereby making spaceflight safer and expanding (...) our knowledge of near-Earth space. To facilitate data-sharing interoperability among distinct orbital debris and space object catalogs, and SSA information systems, I proposed ontology in (Rovetto, 2015) and (Rovetto and Kelso, 2016). I continue this effort toward formal representations and models of the overall domain that may serve to improve peaceful SSA and increase our scientific knowledge. This paper explains the project concept introduced in those publications, summarizing efforts to date as well as the research field of ontology development and engineering. I describe concepts for an ontological framework for the orbital space environment, near-Earth space environment and SSA domain. An ontological framework is conceived as a part of a potential international information system. The purpose of such a system is to consolidate, analyze and reason over various sources and types of orbital and SSA data toward the mutually beneficial goals of safer space navigation and scientific research. Recent internationals findings on the limitations of orbital data, in addition to existing publications on collaborative SSA, demonstrate both the overlap with this project and the need for data-sharing and integration. (shrink)

Space situational awareness (SSA) is vital for international safety and security, and for the future of space travel. The sharing of SSA data and information should improve the state of global SSA for planetary defense and spaceflight safety. I take steps toward a Space Situational Awareness (SSA) Ontology, and outline some central objectives, requirements and desiderata in the ontology development process for this domain. The purpose of this ontological system is to explore the potential for the ontology research topic to (...) (i) represent SSA general knowledge, data, and entities/objects, (ii) clearly express the meaning of SSA data, and (iii) foster SSA data-sharing. The overall goal and motivation is to (iv) improve our capacity for planetary defense, e.g., from near- or deep-space objects and phenomena, and (v) facilitate safer and peaceful space access, navigation and travel, by improving global SSA. This research is thereby intended only for peaceful space-domain applications and uses, with particular interests in orbital debris. There is little application of ontology to the space domain as compared with other disciplines and little if any ontological development of SSA and related domains. In this respect, this paper offers novel concepts. (shrink)

By the end of his life Plato had rearranged the theory of ideas into his teaching about ideal numbers, but no written records have been left. The Ideal mathematics of Plato is present in all his dialogues. It can be clearly grasped in relation to the effective use of mathematical modelling. Many problems of mathematical modelling were laid in the foundation of the method by cutting the three-level idealism of Plato to the single-level “ideism” of Aristotle. For a long time, (...) the real, ideal numbers of Plato’s Ideal mathematics eliminates many mathematical problems, extends the capabilities of modelling, and improves mathematics. (shrink)

Automated reasoning about uncertain knowledge has many applications. One difficulty when developing such systems is the lack of a completely satisfactory integration of logic and probability. We address this problem directly. Expressive languages like higher-order logic are ideally suited for representing and reasoning about structured knowledge. Uncertain knowledge can be modeled by using graded probabilities rather than binary truth-values. The main technical problem studied in this paper is the following: Given a set of sentences, each having some probability of being (...) true, what probability should be ascribed to other (query) sentences? A natural wish-list, among others, is that the probability distribution (i) is consistent with the knowledge base, (ii) allows for a consistent inference procedure and in particular (iii) reduces to deductive logic in the limit of probabilities being 0 and 1, (iv) allows (Bayesian) inductive reasoning and (v) learning in the limit and in particular (vi) allows confirmation of universally quantified hypotheses/sentences. We translate this wish-list into technical requirements for a prior probability and show that probabilities satisfying all our criteria exist. We also give explicit constructions and several general characterizations of probabilities that satisfy some or all of the criteria and various (counter) examples. We also derive necessary and sufficient conditions for extending beliefs about finitely many sentences to suitable probabilities over all sentences, and in particular least dogmatic or least biased ones. We conclude with a brief outlook on how the developed theory might be used and approximated in autonomous reasoning agents. Our theory is a step towards a globally consistent and empirically satisfactory unification of probability and logic. (shrink)

How political opinions are spread on social media has been the subject of many academic researches recently, and rightly so. Social platforms give researchers a unique opportunity to understand how public discourses are perceived, owned and instrumentalized by the general public. This paper is instead focussing on the political discourses themselves, and how a specific machine learning technique - vector space models (VSMs) -, can be used to make systematic and more objective discourse analysis. Political footprints are vector-based representation of (...) a political discourse in which each vector represents a word, they are produced thanks to the training of the English lexicon on large corpora of text. This paper describes a simple implementation of political footprints, some heuristics on how to use them, and their application to four cases: the U.N. Kyoto Protocol and Paris Agreement, the 2008 and 2016 U.S. presidential elections. The reader will be given some reasons to believe that political footprints produce meaningful results, suggestions on how to improve them and validate the results. (shrink)