Latest YouTube Video
Saturday, December 31, 2016
How to create an anonymous function and log a string?
from Google Alert - anonymous http://ift.tt/2hFqmJv
via IFTTT
Not Not Anonymous
from Google Alert - anonymous http://ift.tt/2ioOrB7
via IFTTT
Friday, December 30, 2016
What Youth » @ishodwair filming for Anonymous Zone in Japan. Photo
from Google Alert - anonymous http://ift.tt/2irek5K
via IFTTT
Anonymous Confession
from Google Alert - anonymous http://ift.tt/2hB85v2
via IFTTT
Anonymous Just Hacked Bilderberg & Issued Ominous Threat
from Google Alert - anonymous http://ift.tt/2iNmt1g
via IFTTT
Don't Ban Anonymous Sperm Donations: Study
from Google Alert - anonymous http://ift.tt/2iOcuNe
via IFTTT
Ravens promote WR Keenan Reynolds from practice squad for Sunday's game vs. Bengals; place CB Jimmy Smith (ankle) on IR (ESPN)
via IFTTT
The oddest example of anonymous source reporting that I have ever seen
from Google Alert - anonymous http://ift.tt/2hCptBm
via IFTTT
Re: [FD] [RT-SA-2016-001] Padding Oracle in Apache mod_session_crypto
Source: Gmail -> IFTTT-> Blogger
Obama Expels 35 Russian Spies Over Election Hacking; Russia Responds With Duck Meme
from The Hacker News http://ift.tt/2hyhPq7
via IFTTT
[FD] SwiftMailer <= 5.4.5-DEV Remote Code Execution (CVE-2016-10074)
Source: Gmail -> IFTTT-> Blogger
Re: [FD] [RT-SA-2016-001] Padding Oracle in Apache mod_session_crypto
Source: Gmail -> IFTTT-> Blogger
Thursday, December 29, 2016
Automated timetabling for small colleges and high schools using huge integer programs. (arXiv:1612.08777v1 [cs.AI])
We formulate an integer program to solve a highly constrained academic timetabling problem at the United States Merchant Marine Academy. The IP instance that results from our real case study has approximately both 170,000 rows and columns and solves to near optimality in 12 hours, using a commercial solver. Our model is applicable to both high schools and small colleges who wish to deviate from group scheduling. We also solve a necessary preprocessing student subgrouping problem, which breaks up big groups of students into small groups so they can optimally fit into small capacity classes.
from cs.AI updates on arXiv.org http://ift.tt/2iKl4bw
via IFTTT
The Predictron: End-To-End Learning and Planning. (arXiv:1612.08810v1 [cs.LG])
One of the key challenges of artificial intelligence is to learn models that are effective in the context of planning. In this document we introduce the predictron architecture. The predictron consists of a fully abstract model, represented by a Markov reward process, that can be rolled forward multiple "imagined" planning steps. Each forward pass of the predictron accumulates internal rewards and values over multiple planning depths. The predictron is trained end-to-end so as to make these accumulated values accurately approximate the true value function. We applied the predictron to procedurally generated random mazes and a simulator for the game of pool. The predictron yielded significantly more accurate predictions than conventional deep neural network architectures.
from cs.AI updates on arXiv.org http://ift.tt/2idISaW
via IFTTT
Accelerated Convolutions for Efficient Multi-Scale Time to Contact Computation in Julia. (arXiv:1612.08825v1 [cs.CV])
Convolutions have long been regarded as fundamental to applied mathematics, physics and engineering. Their mathematical elegance allows for common tasks such as numerical differentiation to be computed efficiently on large data sets. Efficient computation of convolutions is critical to artificial intelligence in real-time applications, like machine vision, where convolutions must be continuously and efficiently computed on tens to hundreds of kilobytes per second. In this paper, we explore how convolutions are used in fundamental machine vision applications. We present an accelerated n-dimensional convolution package in the high performance computing language, Julia, and demonstrate its efficacy in solving the time to contact problem for machine vision. Results are measured against synthetically generated videos and quantitatively assessed according to their mean squared error from the ground truth. We achieve over an order of magnitude decrease in compute time and allocated memory for comparable machine vision applications. All code is packaged and integrated into the official Julia Package Manager to be used in various other scenarios.
from cs.AI updates on arXiv.org http://ift.tt/2iKxbFD
via IFTTT
FastMask: Segment Object Multi-scale Candidates in One Shot. (arXiv:1612.08843v1 [cs.CV])
Objects appear to scale differently in natural images. This fact requires methods dealing with object-centric tasks e.g. object proposal to have robust performance over scale variances of objects. In the paper we present a novel segment proposal framework, namely FastMask, which takes advantage of the hierarchical structure in deep convolutional neural network to segment multi-scale objects in one shot. Innovatively, we generalize segment proposal network into three different functional components (body, neck and head). We further propose a weight-shared residual neck module as well as a scale-tolerant attentional head module for multi-scale training and efficient one-shot inference. On MS COCO benchmark, the proposed FastMask outperforms all state-of-the-art segment proposal methods in average recall while keeping 2~5 times faster. More impressively, with a slight trade-off in accuracy, FastMask can segment objects in near real time (~13 fps) at 800$\times$600 resolution images, highlighting its potential in practical applications. Our implementation is available on http://ift.tt/2iKFFAt.
from cs.AI updates on arXiv.org http://ift.tt/2idrHpP
via IFTTT
The formal-logical characterisation of lies, deception, and associated notions. (arXiv:1612.08845v1 [cs.LO])
Defining various dishonest notions in a formal way is a key step to enable intelligent agents to act in untrustworthy environments. This review evaluates the literature for this topic by looking at formal definitions based on modal logic as well as other formal approaches. Criteria from philosophical groundwork is used to assess the definitions for correctness and completeness. The key contribution of this review is to show that only a few definitions fully comply with this gold standard and to point out the missing steps towards a successful application of these definitions in an actual agent environment.
from cs.AI updates on arXiv.org http://ift.tt/2iKuC6h
via IFTTT
Efficient iterative policy optimization. (arXiv:1612.08967v1 [cs.AI])
We tackle the issue of finding a good policy when the number of policy updates is limited. This is done by approximating the expected policy reward as a sequence of concave lower bounds which can be efficiently maximized, drastically reducing the number of policy updates required to achieve good performance. We also extend existing methods to negative rewards, enabling the use of control variates.
from cs.AI updates on arXiv.org http://ift.tt/2idxTOM
via IFTTT
Meta-Unsupervised-Learning: A supervised approach to unsupervised learning. (arXiv:1612.09030v1 [cs.LG])
We introduce a new paradigm to investigate unsupervised learning, reducing unsupervised learning to supervised learning. Specifically, we mitigate the subjectivity in unsupervised decision-making by leveraging knowledge acquired from prior, possibly heterogeneous, supervised learning tasks. We demonstrate the versatility of our framework via comprehensive expositions and detailed experiments on several unsupervised problems such as (a) clustering, (b) outlier detection, and (c) similarity prediction under a common umbrella of meta-unsupervised-learning. We also provide rigorous PAC-agnostic bounds to establish the theoretical foundations of our framework, and show that our framing of meta-clustering circumvents Kleinberg's impossibility theorem for clustering.
from cs.AI updates on arXiv.org http://ift.tt/2iKx9O1
via IFTTT
From Virtual to Real World Visual Perception using Domain Adaptation -- The DPM as Example. (arXiv:1612.09134v1 [cs.CV])
Supervised learning tends to produce more accurate classifiers than unsupervised learning in general. This implies that training data is preferred with annotations. When addressing visual perception challenges, such as localizing certain object classes within an image, the learning of the involved classifiers turns out to be a practical bottleneck. The reason is that, at least, we have to frame object examples with bounding boxes in thousands of images. A priori, the more complex the model is regarding its number of parameters, the more annotated examples are required. This annotation task is performed by human oracles, which ends up in inaccuracies and errors in the annotations (aka ground truth) since the task is inherently very cumbersome and sometimes ambiguous. As an alternative we have pioneered the use of virtual worlds for collecting such annotations automatically and with high precision. However, since the models learned with virtual data must operate in the real world, we still need to perform domain adaptation (DA). In this chapter we revisit the DA of a deformable part-based model (DPM) as an exemplifying case of virtual- to-real-world DA. As a use case, we address the challenge of vehicle detection for driver assistance, using different publicly available virtual-world data. While doing so, we investigate questions such as: how does the domain gap behave due to virtual-vs-real data with respect to dominant object appearance per domain, as well as the role of photo-realism in the virtual world.
from cs.AI updates on arXiv.org http://ift.tt/2iduOxW
via IFTTT
Deep neural heart rate variability analysis. (arXiv:1612.09205v1 [cs.NE])
Despite of the pain and limited accuracy of blood tests for early recognition of cardiovascular disease, they dominate risk screening and triage. On the other hand, heart rate variability is non-invasive and cheap, but not considered accurate enough for clinical practice. Here, we tackle heart beat interval based classification with deep learning. We introduce an end to end differentiable hybrid architecture, consisting of a layer of biological neuron models of cardiac dynamics (modified FitzHugh Nagumo neurons) and several layers of a standard feed-forward neural network. The proposed model is evaluated on ECGs from 474 stable at-risk (coronary artery disease) patients, and 1172 chest pain patients of an emergency department. We show that it can significantly outperform models based on traditional heart rate variability predictors, as well as approaching or in some cases outperforming clinical blood tests, based only on 60 seconds of inter-beat intervals.
from cs.AI updates on arXiv.org http://ift.tt/2iKruYl
via IFTTT
A hybrid approach to supervised machine learning for algorithmic melody composition. (arXiv:1612.09212v1 [cs.AI])
In this work we present an algorithm for composing monophonic melodies similar in style to those of a given, phrase annotated, sample of melodies. For implementation, a hybrid approach incorporating parametric Markov models of higher order and a contour concept of phrases is used. This work is based on the master thesis of Thayabaran Kathiresan (2015). An online listening test conducted shows that enhancing a pure Markov model with musically relevant context, like count and planed melody contour, improves the result significantly.
from cs.AI updates on arXiv.org http://ift.tt/2idwfg0
via IFTTT
Lifted Relational Algebra with Recursion and Connections to Modal Logic. (arXiv:1612.09251v1 [cs.LO])
We propose a new formalism for specifying and reasoning about problems that involve heterogeneous "pieces of information" -- large collections of data, decision procedures of any kind and complexity and connections between them. The essence of our proposal is to lift Codd's relational algebra from operations on relational tables to operations on classes of structures (with recursion), and to add a direction of information propagation. We observe the presence of information propagation in several formalisms for efficient reasoning and use it to express unary negation and operations used in graph databases. We carefully analyze several reasoning tasks and establish a precise connection between a generalized query evaluation and temporal logic model checking. Our development allows us to reveal a general correspondence between classical and modal logics and may shed a new light on the good computational properties of modal logics and related formalisms.
from cs.AI updates on arXiv.org http://ift.tt/2iKkevE
via IFTTT
Attend, Adapt and Transfer: Attentive Deep Architecture for Adaptive Transfer from Multiple Source Tasks. (arXiv:1510.02879v4 [cs.AI] UPDATED)
Transferring knowledge from prior source tasks in solving a new target task can be useful in several learning applications. The application of transfer poses two serious challenges which have not been adequately addressed. First, the agent should be able to avoid negative transfer, which happens when the transfer hampers or slows down the learning instead of helping it. Second, the agent should be able to selectively transfer, which is the ability to select and transfer from different and multiple source tasks for different parts of the state space of the target task. We propose A2T (Attend, Adapt and Transfer), an attentive deep architecture which adapts and transfers from these source tasks. Our model is generic enough to effect transfer of either policies or value functions. Empirical evaluations on different learning algorithms show that A2T is an effective architecture for transfer by being able to avoid negative transfer while transferring selectively from multiple source tasks in the same domain.
from cs.AI updates on arXiv.org http://ift.tt/1MtHoln
via IFTTT
Overcoming Language Variation in Sentiment Analysis with Social Attention. (arXiv:1511.06052v3 [cs.CL] UPDATED)
Variation in language is ubiquitous, particularly in newer forms of writing such as social media. Fortunately, variation is not random; it is often linked to social properties of the author. In this paper, we show how to exploit social networks to make sentiment analysis more robust to social language variation. The key idea is \emph{linguistic homophily}: the tendency of socially linked individuals to use language in similar ways. We formalize this idea in a novel attention-based neural network architecture, in which attention is divided among several basis models, depending on the author's position in the social network. This has the effect of smoothing the classification function across the social network, and makes it possible to induce personalized classifiers even for authors for whom there is no labeled data or demographic metadata. This model significantly improves the accuracies of sentiment analysis on Twitter and review data.
from cs.AI updates on arXiv.org http://ift.tt/1kJ97UX
via IFTTT
Detection of Cooperative Interactions in Logistic Regression Models. (arXiv:1602.03963v2 [cs.AI] UPDATED)
An important problem in the field of bioinformatics is to identify interactive effects among profiled variables for outcome prediction. In this paper, a logistic regression model with pairwise interactions among a set of binary covariates is considered. Modeling the structure of the interactions by a graph, our goal is to recover the interaction graph from independently identically distributed (i.i.d.) samples of the covariates and the outcome.
When viewed as a feature selection problem, a simple quantity called influence is proposed as a measure of the marginal effects of the interaction terms on the outcome. For the case when the underlying interaction graph is known to be acyclic, it is shown that a simple algorithm that is based on a maximum-weight spanning tree with respect to the plug-in estimates of the influences not only has strong theoretical performance guarantees, but can also outperform generic feature selection algorithms for recovering the interaction graph from i.i.d. samples of the covariates and the outcome. Our results can also be extended to the model that includes both individual effects and pairwise interactions via the help of an auxiliary covariate.
from cs.AI updates on arXiv.org http://ift.tt/1WlXJud
via IFTTT
Bank distress in the news: Describing events through deep learning. (arXiv:1603.05670v2 [cs.CL] UPDATED)
While many models are purposed for detecting the occurrence of significant events in financial systems, the task of providing qualitative detail on the developments is not usually as well automated. We present a deep learning approach for detecting relevant discussion in text and extracting natural language descriptions of events. Supervised by only a small set of event information, comprising entity names and dates, the model is leveraged by unsupervised learning of semantic vector representations on extensive text data. We demonstrate applicability to the study of financial risk based on news (6.6M articles), particularly bank distress and government interventions (243 events), where indices can signal the level of bank-stress-related reporting at the entity level, or aggregated at national or European level, while being coupled with explanations. Thus, we exemplify how text, as timely, widely available and descriptive data, can serve as a useful complementary source of information for financial and systemic risk analytics.
from cs.AI updates on arXiv.org http://ift.tt/1o0i4tC
via IFTTT
A Discrete and Bounded Envy-Free Cake Cutting Protocol for Any Number of Agents. (arXiv:1604.03655v11 [cs.DS] UPDATED)
We consider the well-studied cake cutting problem in which the goal is to find an envy-free allocation based on queries from $n$ agents. The problem has received attention in computer science, mathematics, and economics. It has been a major open problem whether there exists a discrete and bounded envy-free protocol. We resolve the problem by proposing a discrete and bounded envy-free protocol for any number of agents. The maximum number of queries required by the protocol is $n^{n^{n^{n^{n^n}}}}$. We additionally show that even if we do not run our protocol to completion, it can find in at most $n^3{(n^2)}^n$ queries a partial allocation of the cake that achieves proportionality (each agent gets at least $1/n$ of the value of the whole cake) and envy-freeness. Finally we show that an envy-free partial allocation can be computed in at most $n^3{(n^2)}^n$ queries such that each agent gets a connected piece that gives the agent at least $1/(3n)$ of the value of the whole cake.
from cs.AI updates on arXiv.org http://ift.tt/1Q85mzr
via IFTTT
Piecewise convexity of artificial neural networks. (arXiv:1607.04917v2 [cs.LG] UPDATED)
Although artificial neural networks have shown great promise in applications including computer vision and speech recognition, there remains considerable practical and theoretical difficulty in optimizing their parameters. The seemingly unreasonable success of gradient descent methods in minimizing these non-convex functions remains poorly understood. In this work we offer some theoretical guarantees for networks with piecewise affine activation functions, which have in recent years become the norm. We prove three main results. Firstly, that the network is piecewise convex as a function of the input data. Secondly, that the network, considered as a function of the parameters in a single layer, all others held constant, is again piecewise convex. Finally, that the network as a function of all its parameters is piecewise multi-convex, a generalization of biconvexity. From here we characterize the local minima and stationary points of the training objective, showing that they minimize certain subsets of the parameter space. We then analyze the performance of two optimization algorithms on multi-convex problems: gradient descent, and a method which repeatedly solves a number of convex sub-problems. We prove necessary convergence conditions for the first algorithm and both necessary and sufficient conditions for the second, after introducing regularization to the objective. Finally, we remark on the remaining difficulty of the global optimization problem. Under the squared error objective, we show that by varying the training data, a single rectifier neuron admits local minima arbitrarily far apart, both in objective value and parameter space.
from cs.AI updates on arXiv.org http://ift.tt/2aoPPjd
via IFTTT
Facebook anonymous proxy
from Google Alert - anonymous http://ift.tt/2id44O2
via IFTTT
ISS Daily Summary Report – 12/28/2016
from ISS On-Orbit Status Report http://ift.tt/2hslgez
via IFTTT
Anonymous Hawker 800XP bizjet from Moscow Vnukovo on a nightly sightseeing tour over Aleppo ...
from Google Alert - anonymous http://ift.tt/2ikpVna
via IFTTT
3 Critical Zero-Day Flaws Found in PHP 7 — One Remains Unpatched!
from The Hacker News http://ift.tt/2hvCN90
via IFTTT
Wednesday, December 28, 2016
Disable cache for blocks for anonymous users
from Google Alert - anonymous http://ift.tt/2hu1Yc3
via IFTTT
Proxy list anonymous l1
from Google Alert - anonymous http://ift.tt/2irQEvp
via IFTTT
New Android Malware Hijacks Router DNS from Smartphone
from The Hacker News http://ift.tt/2i8lMT8
via IFTTT
Anonymous Hawker 800XP bizjet from Moscow Vnukovo on a nightly sightseeing tour over Aleppo ...
from Google Alert - anonymous http://ift.tt/2i8reVK
via IFTTT
Ravens: WR Steve Smith Sr., 37, says he's "89 percent sure" that Sunday will be the final game of his 16-year career (ESPN)
via IFTTT
feathers-authentication-anonymous
from Google Alert - anonymous http://ift.tt/2iEYQMj
via IFTTT
NFL: Ravens WR Steve Smith Sr. to retire after Sunday's game; 5-time Pro Bowler 7th in career receiving yards (14,697) (ESPN)
via IFTTT
FreeBSD Foundation Announces New Uranium Level Donation
from Google Alert - anonymous http://ift.tt/2iERR5T
via IFTTT
ISS Daily Summary Report – 12/27/2016
from ISS On-Orbit Status Report http://ift.tt/2iE4myy
via IFTTT
Police Ask for Amazon Echo Data to Help Solve a Murder Case
from The Hacker News http://ift.tt/2iD4luH
via IFTTT
Anonymous John
from Google Alert - anonymous http://ift.tt/2hvKfCJ
via IFTTT
I have a new follower on Twitter
🔥 Sollevarsi 🔥
Marketing. Social 📈 Content 📚🖼 Web 🖥 Blog 📝 Consulting 📊 #riseup #gritgrow #growthhacking #seo #socialmedia #contentmarketing https://t.co/0oW41NFZeU
Austin, TX
https://t.co/hmOFXu74je
Following: 6195 - Followers: 7188
December 28, 2016 at 04:10AM via Twitter http://twitter.com/riseupsole
Re: [FD] PHPMailer < 5.2.18 Remote Code Execution [CVE-2016-10033]
Source: Gmail -> IFTTT-> Blogger
Re: [FD] [RT-SA-2016-001] Padding Oracle in Apache mod_session_crypto
Source: Gmail -> IFTTT-> Blogger
[FD] PHPMailer < 5.2.20 Remote Code Execution PoC 0day Exploit (CVE-2016-10045) (Bypass of the CVE-2016-1033 patch)
Source: Gmail -> IFTTT-> Blogger
[FD] PHPMailer < 5.2.18 Remote Code Execution [updated advisory] [CVE-2016-10033]
Source: Gmail -> IFTTT-> Blogger
M31: The Andromeda Galaxy
Moon Phase and Libration, 2017 South Up
from NASA's Scientific Visualization Studio: Most Popular
via IFTTT
Tuesday, December 27, 2016
Chameleon in a Candy Store
from Google Alert - anonymous http://ift.tt/2iCIDDi
via IFTTT
Anonymous user 95bbe8
from Google Alert - anonymous http://ift.tt/2i5NtM8
via IFTTT
I have a new follower on Twitter
Expedien, Inc
Data Analytics, Big Data, Data Management, and Business Intelligence services provider, a SAP & Informatica Alliance Partner Firm.
Houston
https://t.co/8QMm2nRRcr
Following: 469 - Followers: 648
December 27, 2016 at 10:48PM via Twitter http://twitter.com/Expedien_Inc
I have a new follower on Twitter
Gina Stepp
Gina Stepp is a writer with a focus on family studies and a master's degree in forensic psychology. Fan of acoustic music--learning to play the mandolin.
Southern California
http://t.co/jq2GM9FHia
Following: 3097 - Followers: 3752
December 27, 2016 at 09:43PM via Twitter http://twitter.com/Gina525
I have a new follower on Twitter
PredictiveHire
#AI for #talent & #HR. Improving productivity, performance, profitability, equality, diversity & social mobility #WorkforceScience #PredictiveAnalytics
Global
https://t.co/PqCDNU0iRj
Following: 7685 - Followers: 9179
December 27, 2016 at 08:48PM via Twitter http://twitter.com/Predictive_Hire
Solving Combinatorial Optimization problems with Quantum inspired Evolutionary Algorithm Tuned using a Novel Heuristic Method. (arXiv:1612.08109v1 [cs.AI])
Quantum inspired Evolutionary Algorithms were proposed more than a decade ago and have been employed for solving a wide range of difficult search and optimization problems. A number of changes have been proposed to improve performance of canonical QEA. However, canonical QEA is one of the few evolutionary algorithms, which uses a search operator with relatively large number of parameters. It is well known that performance of evolutionary algorithms is dependent on specific value of parameters for a given problem. The advantage of having large number of parameters in an operator is that the search process can be made more powerful even with a single operator without requiring a combination of other operators for exploration and exploitation. However, the tuning of operators with large number of parameters is complex and computationally expensive. This paper proposes a novel heuristic method for tuning parameters of canonical QEA. The tuned QEA outperforms canonical QEA on a class of discrete combinatorial optimization problems which, validates the design of the proposed parameter tuning framework. The proposed framework can be used for tuning other algorithms with both large and small number of tunable parameters.
from cs.AI updates on arXiv.org http://ift.tt/2hLOBBv
via IFTTT
Theory-guided Data Science: A New Paradigm for Scientific Discovery. (arXiv:1612.08544v1 [cs.LG])
Data science models, although successful in a number of commercial domains, have had limited applicability in scientific problems involving complex physical phenomena. Theory-guided data science (TGDS) is an emerging paradigm that aims to leverage the wealth of scientific knowledge for improving the effectiveness of data science models in enabling scientific discovery. The overarching vision of TGDS is to introduce scientific consistency as an essential component for learning generalizable models. Further, by producing scientifically interpretable models, TGDS aims to advance our scientific understanding by discovering novel domain insights. Indeed, the paradigm of TGDS has started to gain prominence in a number of scientific disciplines such as turbulence modeling, material discovery, quantum chemistry, bio-medical science, bio-marker discovery, climate science, and hydrology. In this paper, we formally conceptualize the paradigm of TGDS and present a taxonomy of research themes in TGDS. We describe several approaches for integrating domain knowledge in different research themes using illustrative examples from different disciplines. We also highlight some of the promising avenues of novel research for realizing the full potential of theory-guided data science.
from cs.AI updates on arXiv.org http://ift.tt/2huWMpL
via IFTTT
Monte Carlo Sort for unreliable human comparisons. (arXiv:1612.08555v1 [cs.AI])
Algorithms which sort lists of real numbers into ascending order have been studied for decades. They are typically based on a series of pairwise comparisons and run entirely on chip. However people routinely sort lists which depend on subjective or complex judgements that cannot be automated. Examples include marketing research; where surveys are used to learn about customer preferences for products, the recruiting process; where interviewers attempt to rank potential employees, and sporting tournaments; where we infer team rankings from a series of one on one matches. We develop a novel sorting algorithm, where each pairwise comparison reflects a subjective human judgement about which element is bigger or better. We introduce a finite and large error rate to each judgement, and we take the cost of each comparison to significantly exceed the cost of other computational steps. The algorithm must request the most informative sequence of comparisons from the user; in order to identify the correct sorted list with minimum human input. Our Discrete Adiabatic Monte Carlo approach exploits the gradual acquisition of information by tracking a set of plausible hypotheses which are updated after each additional comparison.
from cs.AI updates on arXiv.org http://ift.tt/2hLQJcw
via IFTTT
A Sparse Nonlinear Classifier Design Using AUC Optimization. (arXiv:1612.08633v1 [cs.AI])
AUC (Area under the ROC curve) is an important performance measure for applications where the data is highly imbalanced. Learning to maximize AUC performance is thus an important research problem. Using a max-margin based surrogate loss function, AUC optimization problem can be approximated as a pairwise rankSVM learning problem. Batch learning methods for solving the kernelized version of this problem suffer from scalability and may not result in sparse classifiers. Recent years have witnessed an increased interest in the development of online or single-pass online learning algorithms that design a classifier by maximizing the AUC performance. The AUC performance of nonlinear classifiers, designed using online methods, is not comparable with that of nonlinear classifiers designed using batch learning algorithms on many real-world datasets. Motivated by these observations, we design a scalable algorithm for maximizing AUC performance by greedily adding the required number of basis functions into the classifier model. The resulting sparse classifiers perform faster inference. Our experimental results show that the level of sparsity achievable can be order of magnitude smaller than the Kernel RankSVM model without affecting the AUC performance much.
from cs.AI updates on arXiv.org http://ift.tt/2huZ3Be
via IFTTT
Role of Simplicity in Creative Behaviour: The Case of the Poietic Generator. (arXiv:1612.08657v1 [cs.AI])
We propose to apply Simplicity Theory (ST) to model interest in creative situations. ST has been designed to describe and predict interest in communication. Here we use ST to derive a decision rule that we apply to a simplified version of a creative game, the Poietic Generator. The decision rule produces what can be regarded as an elementary form of creativity. This study is meant as a proof of principle. It suggests that some creative actions may be motivated by the search for unexpected simplicity.
from cs.AI updates on arXiv.org http://ift.tt/2hv5kNt
via IFTTT
The Linearization of Belief Propagation on Pairwise Markov Networks. (arXiv:1502.04956v2 [cs.AI] UPDATED)
Belief Propagation (BP) is a widely used approximation for exact probabilistic inference in graphical models, such as Markov Random Fields (MRFs). In graphs with cycles, however, no exact convergence guarantees for BP are known, in general. For the case when all edges in the MRF carry the same symmetric, doubly stochastic potential, recent works have proposed to approximate BP by linearizing the update equations around default values, which was shown to work well for the problem of node classification. The present paper generalizes all prior work and derives an approach that approximates loopy BP on any pairwise MRF with the problem of solving a linear equation system. This approach combines exact convergence guarantees and a fast matrix implementation with the ability to model heterogenous networks. Experiments on synthetic graphs with planted edge potentials show that the linearization has comparable labeling accuracy as BP for graphs with weak potentials, while speeding-up inference by orders of magnitude.
from cs.AI updates on arXiv.org http://ift.tt/1MvHhEQ
via IFTTT
Thailand Censorship OpSingleGateway
from Google Alert - anonymous http://ift.tt/2hLzPe0
via IFTTT
Ravens: John Harbaugh won't be entering final year of deal despite speculation he was only signed through 2017 - Hensley (ESPN)
via IFTTT
I have a new follower on Twitter
Wings DAO Platform
Wings is a cross blockchain Decentralized Autonomous Organizations management platform that allows easy DAO setup, participation or administration.
https://t.co/tLUO7BVy8C
Following: 4530 - Followers: 7916
December 27, 2016 at 03:48PM via Twitter http://twitter.com/wingsplatform
Anonymous Hacks and Defaces Thai LA Consulate to Protest Arrests and Cyber Law
from Google Alert - anonymous http://ift.tt/2iANwkl
via IFTTT
Anonymous' recently released journalist Barret Brown still vows never to back down
from Google Alert - anonymous http://ift.tt/2inOFIB
via IFTTT
Did You Install Super Mario Run APK for Android? That's Malware
from The Hacker News http://ift.tt/2htJvxG
via IFTTT
I have a new follower on Twitter
Harun Tahta
Tahta tabi, zoruna mı gitti?
Following: 21 - Followers: 15
December 27, 2016 at 09:48AM via Twitter http://twitter.com/TahtaHarun
ISS Daily Summary Report – 12/23/2016
from ISS On-Orbit Status Report http://ift.tt/2hkg3FI
via IFTTT
ISS Daily Summary Report – 12/22/2016
from ISS On-Orbit Status Report http://ift.tt/2hpEO6C
via IFTTT
I have a new follower on Twitter
Haşim İşgüzar
Tembel İnsan
İstanbul
Following: 23 - Followers: 6
December 27, 2016 at 05:48AM via Twitter http://twitter.com/HasimIsG
Re: [FD] [RT-SA-2016-001] Padding Oracle in Apache mod_session_crypto
Source: Gmail -> IFTTT-> Blogger
[FD] PHPMailer < 5.2.18 Remote Code Execution [CVE-2016-10033]
Source: Gmail -> IFTTT-> Blogger
[FD] kernel vuln status question - how can I be protected
Source: Gmail -> IFTTT-> Blogger
[FD] Arbitrary file deletion vulnerability in Image Slider allows authenticated users to delete files (WordPress plugin)
Source: Gmail -> IFTTT-> Blogger
Stabat Sancta Maria (Anonymous)
from Google Alert - anonymous http://ift.tt/2hJiMJR
via IFTTT
Adoramus Te Christe (Anonymous)
from Google Alert - anonymous http://ift.tt/2ic2qwu
via IFTTT
Monday, December 26, 2016
Alcoholics Anonymous - Big Book
from Google Alert - anonymous http://ift.tt/2i92FWW
via IFTTT
I have a new follower on Twitter
Kaseya Corp
Kaseya solutions allow MSPs and IT organizations to efficiently manage and secure IT in order to drive IT service and business success.
Boston, MA, USA
http://t.co/rbftW6eGgW
Following: 6254 - Followers: 9689
December 26, 2016 at 07:13PM via Twitter http://twitter.com/KaseyaCorp
Sundays and Wednesday evenings
from Google Alert - anonymous http://ift.tt/2ixIWim
via IFTTT
What is the difference between Normal function vs Anonymous Function?
from Google Alert - anonymous http://ift.tt/2hhfnAY
via IFTTT
Anonymous' Journalist Released after 4 Years in Jail -- Vows Never to Back Down
from Google Alert - anonymous http://ift.tt/2hh9HGP
via IFTTT
OpenCV: Resolving NoneType errors
Each week I receive and respond to at least 2-3 emails and 3-4 blog post comments regarding
NoneTypeerrors in OpenCV and Python.
For beginners, these errors can be hard to diagnose — by definition they aren’t very informative.
Since this question is getting asked so often I decided to dedicate an entire blog post to the topic.
While
NoneTypeerrors can be caused for a nearly unlimited number of reasons, in my experience, both as a computer vision developer and chatting with other programmers here on PyImageSearch, in over 95% of the cases,
NoneTypeerrors in OpenCV are caused by either:
- An invalid image path passed to
cv2.imread
. - A problem reading a frame from a video stream/video file via
cv2.VideoCapture
and the associated.read
method.
To learn more about
NoneTypeerrors in OpenCV (and how to avoid them), just keep reading.
Looking for the source code to this post?
Jump right to the downloads section.
OpenCV: Resolving NoneType errors
In the first part of this blog post I’ll discuss exactly what
NoneTypeerrors are in the Python programming language.
I’ll then discuss the two primary reasons you’ll run into
NoneTypeerrors when using OpenCV and Python together.
Finally, I’ll put together an actual example that not only causes a
NoneTypeerror, but also resolves it as well.
What is a NoneType error?
When using the Python programming language you’ll inevitably run into an error that looks like this:
AttributeError: 'NoneType' object has no attribute ‘something’
Where
somethingcan be replaced by whatever the name of the actual attribute is.
We see these errors when we think we are working with an instance of a particular Class or Object, but in reality we have the Python built-in type
None.
As the name suggests,
Nonerepresents the absence of a value, such as when a function call returns an unexpected result or fails entirely.
Here is an example of generating a
NoneTypeerror from the Python shell:
>>> foo = None >>> foo.bar = True Traceback (most recent call last): File "<stdin>", line 1, in <module> AttributeError: 'NoneType' object has no attribute 'bar' >>>
Here I create a variable named
fooand set it to
None.
I then try to set the
barattribute of
footo
True, but since
foois a
NoneTypeobject, Python will not allow this — hence the error message.
Two reasons for 95% of OpenCV NoneType errors
When using OpenCV and Python bindings, you’re bound to come across
NoneTypeerrors at some point.
In my experience, over 95% of the time these
NoneTypeerrors can be traced back to either an issue with
cv2.imreador
cv2.VideoCapture.
I have provided details for each of the cases below.
Case #1: cv2.imread
If you are receiving a
NoneTypeerror and your code is calling
cv2.imread, then the likely cause of the error is an invalid file path supplied to
cv2.imread.
The
cv2.imreadfunction does not explicitly throw an error message if you give it an invalid file path (i.e., a path to a nonexistent file). Instead,
cv2.imreadwill simply return
None.
Anytime you try to access an attribute of a
Noneimage loaded from disk via
cv2.imreadyou’ll get a
NoneTypeerror.
Here is an example of trying to load a nonexistent image from disk:
$ python >>> import cv2 >>> path = "path/to/image/that/does/not/exist.png" >>> image = cv2.imread(path) >>> print(image.shape) Traceback (most recent call last): File "<stdin>", line 1, in <module> AttributeError: 'NoneType' object has no attribute 'shape'
As you can see,
cv2.imreadgladly accepts the image path (even though it doesn’t exist), realizes the image path is invalid, and returns
None. This is especially confusing for Python programmers who are used to these types of functions throwing exceptions.
As an added bonus, I’ll also mention the
AssertionFailedexception.
If you try to pass an invalid image (i.e.,
NoneTypeimage) into another OpenCV function, Python + OpenCV will complain that the image doesn’t have any width, height, or depth information — and how could it, the “image” is a
Noneobject after all!
Here is an example of an error message you might see when loading a nonexistent image from disk and followed by immediately calling an OpenCV function on it:
>>> import cv2 >>> path = "path/to/image/that/does/not/exist.png" >>> image = cv2.imread(path) >>> gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) OpenCV Error: Assertion failed (scn == 3 || scn == 4) in cvtColor, file /tmp/opencv20150906-42178-3d0iam/opencv-2.4.12/modules/imgproc/src/color.cpp, line 3739 Traceback (most recent call last): File "<stdin>", line 1, in <module> cv2.error: /tmp/opencv20150906-42178-3d0iam/opencv-2.4.12/modules/imgproc/src/color.cpp:3739: error: (-215) scn == 3 || scn == 4 in function cvtColor >>>
These types of errors can be harder to debug since there are many reasons why an
AssertionErrorcould be thrown. But in most cases, your first step should be be ensuring that your image was correctly loaded from disk.
A final, more rare problem you may encounter with
cv2.imreadis that your image does exist on disk, but you didn’t compile OpenCV with the given image I/O libraries installed.
For example, let’s say you have a .JPEG file on disk and you knew you had the correct path to it.
You then try to load the JPEG file via
cv2.imreadand notice a
NoneTypeor
AssertionError.
How can this be?
The file exists!
In this case, you likely forgot to compile OpenCV with JPEG file support enabled.
In Debian/Ubuntu systems, this is caused by a lack of
libjpegbeing installed.
For macOS systems, you likely forgot to install the
jpeglibrary via Homebrew.
To resolve this problem, regardless of operating system, you’ll need to re-compile and re-install OpenCV. Please see this page for more details on how to compile and install OpenCV on your particular system.
Case #2: cv2.VideoCapture and .read
Just like we see
NoneTypeerrors and
AssertionErrorexceptions when using
cv2.imread, you’ll also see these errors when working with video streams/video files as well.
To access a video stream, OpenCV uses the
cv2.VideoCapturewhich accepts a single argument, either:
- A string representing the path to a video file on disk.
- An integer representing the index of a webcam on your computer.
Working with video streams and video files with OpenCV is more complex than simply loading an image via
cv2.imread, but the same rules apply.
If you try to call the
.readmethod of an instantiated
cv2.VideoCapture(regardless if it’s a video file or webcam stream) and notice a
NoneTypeerror or
AssertionError, then you likely have a problem with either:
- The path to your input video file (it’s probably incorrect).
- Not having the proper video codecs installed, in which case you’ll need to install the codecs, followed by re-compiling and re-installing OpenCV (see this page for a complete list of tutorials).
- Your webcam not being accessible via OpenCV. This could be for any number of reasons, including missing drivers, an incorrect index passed to
cv2.VideoCapture
, or simply your webcam is not properly attached to your system.
Again, working with video files is more complex than working with simple image files, so make sure you’re systematic in resolving the issue.
First, try to access your webcam via a separate piece of software than OpenCV.
Or, try to load your video file in a movie player.
If both of those work, you likely have a problem with your OpenCV install.
Otherwise, it’s most likely a codec or driver issue.
An example of creating and resolving an OpenCV NoneType error
To demonstrate a
NoneTypeerror in action I decided to create a highly simplified Python + OpenCV script that represents what you might see elsewhere on the PyImageSearch blog.
Open up a new file, name it
display_image.py, and insert the following code:
# import the necessary packages import argparse import cv2 # construct the argument parse and parse the arguments ap = argparse.ArgumentParser() ap.add_argument("-i", "--image", required=True, help="path to the image file") args = vars(ap.parse_args()) # load the image from disk and display the width, height, # and depth image = cv2.imread(args["image"]) (h, w, d) = image.shape print("w: {}, h: {}, d: {}".format(w, h, d)) # show the image cv2.imshow("Image", image) cv2.waitKey(0)
All this script does is:
- Parse command line arguments.
- (Attempts to) load an image from disk.
- Prints the width, height, and depth of the image to the terminal.
- Displays the image to our screen.
For most Python developers who are familiar with the command line, this script won’t give you any trouble.
But if you’re new to the command line and are unfamiliar/uncomfortable with command line arguments, you can easily run into a
NoneTypeerror if you’re not careful.
How, you might say?
The answer lies in not properly using/understanding command line arguments.
Over the past few years of running this blog, I’ve seen many emails and blog post comments from readers who are trying to modify the
.add_argumentfunction to supply the path to their image file.
DON’T DO THIS — you don’t have to change a single line of argument parsing code.
Instead, what you should do is spend the next 10 minutes reading through this excellent article that explains what command line arguments are and how to use them in Python:
This is required reading if you expect to follow tutorials here on the PyImageSearch blog.
Working with the command line, and therefore command line arguments, are a big part of what it means to be a computer scientist — a lack of command line skills is only going to harm you. You’ll thank me later.
Going back to the example, let’s check the contents of my local directory:
$ ls -l total 800 -rw-r--r-- 1 adrianrosebrock staff 541 Dec 21 08:45 display_image.py -rw-r--r-- 1 adrianrosebrock staff 403494 Dec 21 08:45 jemma.png
As we can see, I have two files:
-
display_image.py
: My Python script that I’ll be executing shortly. -
jemma.png
: The photo I’ll be loading from disk.
If I execute the following command I’ll see the
jemma.pngimage displayed to my screen, along with information on the dimensions of the image:
$ python display_image.py --image jemma.png w: 376, h: 500, d: 3
However, let’s try to load an image path that does not exist:
$ python display_image.py --image i_dont_exist.png Traceback (most recent call last): File "display_image.py", line 17, in <module> (h, w, d) = image.shape AttributeError: 'NoneType' object has no attribute 'shape'
Sure enough, there is our
NoneTypeerror.
In this case, it was caused because I did not supply a valid image path to
cv2.imread.
Summary
In this blog post I discussed
NoneTypeerrors and
AssertionErrorexceptions in OpenCV and Python.
In the vast majority of these situations, these errors can be attributed to either the
cv2.imreador
cv2.VideoCapturemethods.
Whenever you encounter one of these errors, make sure you can load your image/read your frame before continuing. In over 95% of circumstances, your image/frame was not properly read.
Otherwise, if you are using command line arguments and are unfamiliar with them, there is a chance that you aren’t using them properly. In that case, make sure you educate yourself by reading this tutorial on command line arguments — you’ll thank me later.
Anyway, I hope this tutorial has helped you in your journey to OpenCV mastery!
If you’re just getting started studying computer vision and OpenCV, I would highly encourage you to take a look at my book, Practical Python and OpenCV, which will help you grasp the fundamentals.
Otherwise, make sure you enter your email address in the form below to be notified when future blog posts and tutorials are published!
Downloads:
The post OpenCV: Resolving NoneType errors appeared first on PyImageSearch.
from PyImageSearch http://ift.tt/2hr35us
via IFTTT