09
Gen
21

Explaining and Harnessing Adversarial Examples

Several machine learning models, including neural networks, consistently misclassify adversarial examples—inputs formed by applying small but intentionally worst-case perturbations to examples from the dataset, such that the perturbed input results in the model outputting an incorrect answer with high confidence. Early attempts at explaining this phenomenon focused on nonlinearity and overfitting. We argue instead that the primary cause of neural networks’ vulnerability to adversarial perturbation is their linear nature. This explanation is supported by new quantitative results while giving the first explanation of the most intriguing fact about them: their generalization across architectures and training sets. Moreover, this view yields a simple and fast method of generating adversarial examples. Using this approach to provide examples for adversarial training, we reduce the test set error of a maxout network on the MNIST dataset.

https://arxiv.org/abs/1412.6572

09
Gen
21

Adversarial examples in the physical world

Most existing machine learning classifiers are highly vulnerable to adversarial examples. An adversarial example is a sample of input data which has been modified very slightly in a way that is intended to cause a machine learning classifier to misclassify it. In many cases, these modifications can be so subtle that a human observer does not even notice the modification at all, yet the classifier still makes a mistake. Adversarial examples pose security concerns because they could be used to perform an attack on machine learning systems, even if the adversary has no access to the underlying model. Up to now, all previous work have assumed a threat model in which the adversary can feed data directly into the machine learning classifier. This is not always the case for systems operating in the physical world, for example those which are using signals from cameras and other sensors as an input. This paper shows that even in such physical world scenarios, machine learning systems are vulnerable to adversarial examples. We demonstrate this by feeding adversarial images obtained from cell-phone camera to an ImageNet Inception classifier and measuring the classification accuracy of the system. We find that a large fraction of adversarial examples are classified incorrectly even when perceived through the camera.

Nessuna descrizione alternativa per questa immagine

https://arxiv.org/abs/1607.02533

09
Gen
21

Mastering the game of Go with deep neural networks and tree search

Abstract

The game of Go has long been viewed as the most challenging of classic games for artificial intelligence owing to its enormous search space and the difficulty of evaluating board positions and moves. Here we introduce a new approach to computer Go that uses ‘value networks’ to evaluate board positions and ‘policy networks’ to select moves. These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play. Without any lookahead search, the neural networks play Go at the level of state-of-the-art Monte Carlo tree search programs that simulate thousands of random games of self-play. We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks. Using this search algorithm, our program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0. This is the first time that a computer program has defeated a human professional player in the full-sized game of Go, a feat previously thought to be at least a decade away.

https://www.nature.com/articles/nature16961

11
Dic
20

La leadership “agile”, incontro con il Generale di Divisione Aerea Fernando Giancotti

Fernando Giancotti e Yakov Shaharabani, Leadership agile nella complessità. Organizzazioni, Stormi da combattimento, Guerini e Associati 2008

Nell’ultimo “snodo organizzativo”, intitolato “Agilità e antifragilità: le società ipermoderne possono apprendere dalla culture arcaiche?” abbiamo conosciuto i best seller internazionali dell’antropologo Diamond e dello statistico Taleb che ci invitano a prendere coscienza e ad approfittare – anche nel business management – di un’originaria saggezza derivante dall’adattamento del nostro “hardware” alle condizioni difficili della vita primitiva. Su quell’arcaica, preistorica saggezza “cablata” in noi, frutto del lento meccanismo di selezione naturale, si è stratificata una nuova saggezza di tipo culturale: un software continuamente perfezionato, frutto della creazione di mezzi di gestione, conservazione e trasmissione della conoscenza come il linguaggio prima, la scrittura decine di migliaia di anni dopo, la stampa mezzo millennio fa e infine le tecnologie e i media elettronici analogici e digitali. Nel frattempo, mentre avveniva tutto questo (nel corso di quel breve periodo che chiamiamo storia), il mondo diventava quello che conosciamo adesso, che tutti riconoscono essere molto complesso. Una società in cui ci muoviamo tra tentativi ed errori,  ancora fortemente guidati da routines del nostro preistorico e collaudato firmware. Capire ed assecondare tali arcaiche routines può essere per il manager un punto di forza.

https://viceconte.wordpress.com/2013/12/16/1070/

Diapositiva2
11
Dic
20

Meccanismi molecolari del rilevamento della gravità delle radici e della trasduzione del segnale

Le piante usano la gravità come guida per dirigere le loro radici nel terreno per ancorarsi e per trovare le risorse necessarie per la crescita e lo sviluppo. Nelle piante superiori, le cellule di columella della punta della radice formano il sito primario di rilevamento della gravità, e in queste cellule la sedimentazione di plastidi densi e ricchi di amido (amiloplasti) innesca la trasduzione del segnale di gravità. Questo genera un gradiente di auxina attraverso il cappello della radice che viene trasmesso alla zona di allungamento dove promuove l’allungamento cellulare differenziale, consentendo alla radice di dirigersi verso il basso. Non è ancora ben compreso come la sedimentazione dell’amiloplasto porti alla ridistribuzione dell’auxina. Sono stati proposti modelli per spiegare come i canali ionici meccanosensibili o le interazioni ligando-recettore potrebbero collegare questi eventi. Sebbene i loro ruoli non siano ancora chiari, possibili secondi messaggeri in questo processo includono protoni, Ca (2+) e inositolo 1,4,5-trifosfato. Dopo gravistimolazione, i facilitatori di efflusso di auxina PIN3 e PIN7 si riposizionano sul lato inferiore delle cellule della columella e mediano la ridistribuzione dell’auxina. Tuttavia, l’evidenza di un meccanismo secondario indipendente dall’auxina di rilevamento della gravità e trasduzione del segnale suggerisce che questo processo fisiologico è piuttosto complesso. Inoltre, le piante devono integrare una varietà di segnali ambientali, risultando in relazioni sfaccettate tra gravitropismo e altre risposte di crescita direzionale come idro, foto e thigmotropismo. i facilitatori dell’efflusso di auxina PIN3 e PIN7 si riposizionano sul lato inferiore delle cellule della columella e mediano la ridistribuzione dell’auxina. Tuttavia, l’evidenza di un meccanismo secondario indipendente dall’auxina di rilevamento della gravità e trasduzione del segnale suggerisce che questo processo fisiologico è piuttosto complesso. Inoltre, le piante devono integrare una varietà di segnali ambientali, risultando in relazioni sfaccettate tra gravitropismo e altre risposte di crescita direzionale come idro, foto e thigmotropismo. i facilitatori dell’efflusso di auxina PIN3 e PIN7 si riposizionano sul lato inferiore delle cellule della columella e mediano la ridistribuzione dell’auxina. Tuttavia, l’evidenza di un meccanismo secondario indipendente dall’auxina di rilevamento della gravità e trasduzione del segnale suggerisce che questo processo fisiologico è piuttosto complesso. Inoltre, le piante devono integrare una varietà di segnali ambientali, risultando in relazioni sfaccettate tra gravitropismo e altre risposte di crescita direzionale come idro, foto e thigmotropismo.

https://pubmed.ncbi.nlm.nih.gov/23801441/

Cos'è il geotropismo o il gravitropismo? - 2020
14
Nov
20

The Extended Theory of Cognitive Creativity

This edited volume focuses on the hypothesis that performativity is not a property confined to certain specific human skills, or to certain specific acts of language, nor an accidental enrichment due to creative intelligence. Instead, the executive and motor component of cognitive behavior should be considered an intrinsic part of the physiological functioning of the mind, and as endowed with self-generative power. Performativity, in this theoretical context, can be defined as a constituent component of cognitive processes. The material action allowing us to interact with reality is both the means by which the subject knows the surrounding world and one through which he experiments with the possibilities of his body. This proposal is rooted in models now widely accepted in the philosophy of mind and language; in fact, it focuses on a space of awareness that is not in the individual, or outside it, but is determined by the species-specific ways in which the body acts on the world. This theoretical hypothesis will be pursued through the latest interdisciplinary methodology typical of cognitive science, that coincide with the five sections in which the book is organized: Embodied, enactivist, philosophical approaches; Aesthetics approaches; Naturalistic and evolutionary approaches; Neuroscientific approaches; Linguistics approaches. This book is intended for: linguists, philosophers, psychologists, cognitive scientists, scholars of art and aesthetics, performing artists, researchers in embodied cognition, especially enactivists and students of the extended mind.

https://drive.google.com/file/d/1Om1x7E8FsHzMvmtJZ2D_C4u4cgCb6DOT/view

14
Nov
20

Reimagining ethical leadershipas a relational, contextual andpolitical practice

Helena Liu
Swinburne University of Technology, Victoria, Australia

Abstract

Interest in ethical leadership has been spurred by the widespread reporting of corporate malfeasance and corruption in the last decade. Although ethical leadership theories have highlighted the importance of ethical considerations in leadership, the dominant discourses of this field tend to treat ethical leadership as individualised, decontextualised and power-neutral. The purpose of this article is to address these limitations of the mainstream literature through a reimagination of ethical leadership research, development and practice grounded in a feminist, communitarian and corporeal ethic. This approach, I propose, has the potential to reorient
leadership as a collective ethico-political project exercised towards the goals of equality, justice and emancipation.

https://drive.google.com/file/d/0Bz8cVS8LoO7OZGNSVzMzdXdEdEk/view

14
Nov
20

Ethical Leadership: A review and future directions

Our literature review focuses on the emerging construct of ethical leadership and compares this construct with related concepts that share a common concern for a moral dimension of leadership (e.g., spiritual, authentic, and transformational leadership). Drawing broadly from the intersection of the ethics and leadership literature, we offer propositions about the antecedents and outcomes of ethical leadership. We also identify issues and questions to be addressed in the future and discuss their implications for research and practice. Our review indicates that ethical leadership remains largely unexplored, offering researchers opportunities for new discoveries and leaders opportunities to improve their effectiveness.

Ethical Leadership: A review and future directions

08
Nov
20

Risk Orientations and Policy Frames

Abstract

In this article, we examine the effect of citizens’ risk orientations on policy choices that are framed in various ways. We introduce an original risk orientations scale and test for the relationship between risk orientations and policy preferences using an original survey experiment. We find that individuals with higher levels of risk acceptance are more likely to prefer probabilistic outcomes as opposed to certain outcomes. Mortality and survival frames influence the choices citizens make, but so does our individual-difference measure of risk acceptance. Finally, using a unique within-subject design, we find that risk acceptance undercuts susceptibility to framing effects across successive framing scenarios. The findings suggest that citizens’ risk orientations are consequential in determining their policy views and their susceptibility to framing effects.

https://www.journals.uchicago.edu/doi/abs/10.1017/S0022381609990806

PDF] Risk Orientation, Risk Exposure, and Policy Opinions: The Case of Free  Trade | Semantic Scholar
08
Nov
20

Neurocognitive correlates of liberalism and conservatism

David M AmodioJohn T JostSarah L Master & Cindy M Yee 

Abstract

Political scientists and psychologists have noted that, on average, conservatives show more structured and persistent cognitive styles, whereas liberals are more responsive to informational complexity, ambiguity and novelty. We tested the hypothesis that these profiles relate to differences in general neurocognitive functioning using event-related potentials, and found that greater liberalism was associated with stronger conflict-related anterior cingulate activity, suggesting greater neurocognitive sensitivity to cues for altering a habitual response pattern.

https://www.nature.com/articles/nn1979




Time is real? I think not

gennaio: 2022
L M M G V S D
 12
3456789
10111213141516
17181920212223
24252627282930
31  

Commenti recenti

Inserisci il tuo indirizzo e-mail per iscriverti a questo blog e ricevere notifiche di nuovi messaggi per e-mail.

Unisciti ad altri 1.145 follower

Latest Tweets


%d blogger hanno fatto clic su Mi Piace per questo: