The Internet is full of fun fake images -- from flying sharks and cows on cars to a dizzying variety of celebrity mashups. Hyperrealistic image and video fakes generated by convolutional neural networks (CNNs) however are no laughing matter -- in fact they can be downright dangerous. Deepfake porn reared its ugly head in 2018, fake political speeches by world leaders have cast doubt on news sources, and during the recent Australian bushfires manipulated images mislead people regarding the location and size of fires. Fake images and videos are giving AI a black eye -- but how can the machine learning community fight back? A new paper from UC Berkeley and Adobe researchers declares war on fake images.

Police in London are moving ahead with a deploying a facial recognition camera system despite privacy concerns and evidence that the technology is riddled with false positives. The Metropolitan Police, the U.K.'s biggest police department with jurisdiction over most of London, announced Friday it would begin rolling out new "live facial recognition" cameras in London, making the capital one of the largest cities in the West to adopt the controversial technology. The "Met," as the police department is known in London, said in a statement the facial recognition technology, which is meant to identify people on a watch list and alert police to their real-time location, would be "intelligence-led" and deployed to only specific locations. It's expected to be rolled out as soon as next month. However, privacy activists immediately raised concerns, noting that independent reviews of trials of the technology showed a failure rate of 81%.

When it comes to trade surveillance, regulators want firms to do their own alert calibration, examine all alerts, and keep auditable records. Firms need to balance the real cost of false positives with the technical challenge and risk of self-calibrating and auto-calibrating, while compliance, IT and vendors have to grapple with the need for defensible and transparent audit, which challenges dynamic parameters. The webinar will review recent regulatory statements noting concerns about how trading organisations are setting parameters and managing surveillance. Moving on, it will discuss approaches and technologies that can mitigate these concerns, and question whether advanced approaches such as machine learning are a help or hindrance. Finally, it will set out practical plans for achieving successful surveillance for Market Abuse Regulation (MAR).

Voice-triggered smart assistants often rely on detection of a trigger-phrase before they start listening for the user request. Mitigation of false triggers is an important aspect of building a privacy-centric non-intrusive smart assistant. In this paper, we address the task of false trigger mitigation (FTM) using a novel approach based on analyzing automatic speech recognition (ASR) lattices using graph neural networks (GNN). The proposed approach uses the fact that decoding lattice of a falsely triggered audio exhibits uncertainties in terms of many alternative paths and unexpected words on the lattice arcs as compared to the lattice of a correctly triggered audio. A pure trigger-phrase detector model doesn't fully utilize the intent of the user speech whereas by using the complete decoding lattice of user audio, we can effectively mitigate speech not intended for the smart assistant. We deploy two variants of GNNs in this paper based on 1) graph convolution layers and 2) self-attention mechanism respectively. Our experiments demonstrate that GNNs are highly accurate in FTM task by mitigating ~87% of false triggers at 99% true positive rate (TPR). Furthermore, the proposed models are fast to train and efficient in parameter requirements.

Outlier detection has received special attention in various fields, mainly for those dealing with machine learning and artificial intelligence. As strong outliers, anomalies are divided into the point, contextual and collective outliers. The most important challenges in outlier detection include the thin boundary between the remote points and natural area, the tendency of new data and noise to mimic the real data, unlabelled datasets and different definitions for outliers in different applications. Considering the stated challenges, we defined new types of anomalies called Collective Normal Anomaly and Collective Point Anomaly in order to improve a much better detection of the thin boundary between different types of anomalies. Basic domain-independent methods are introduced to detect these defined anomalies in both unsupervised and supervised datasets. The Multi-Layer Perceptron Neural Network is enhanced using the Genetic Algorithm to detect newly defined anomalies with higher precision so as to ensure a test error less than that calculated for the conventional Multi-Layer Perceptron Neural Network. Experimental results on benchmark datasets indicated reduced error of anomaly detection process in comparison to baselines.

How can we ensure that a machine learning system produces an o utput that generalizes to the underlying distribution, rather than overfitting its train ing data? That is, how can we ensure that the hypotheses or models that are produced are reflective of t he underlying population the training data was drawn from, rather than patterns that occur only by c hance in the training data? This is perhaps the fundamental question for the science of statist ical machine learning. A vast array of methods have been proposed to answer this ques tion. Most notably, the theory of uniform convergence shows that, if the output is sufficiently "simple," then it cannot overfit too much. A more recent line of work has used distributional stability (in the form of differential privacy) to provide generalization guarantees that compose adaptivel y - that is, statistical validity is preserved even when a dataset is reused multiple times with each succes sive analysis being influenced by the outcomes of prior analyses. Other methods for proving gener alization include compression schemes and uniform stability. Unfortunately, these different methods for providing gener alization guarantees are largely disconnected from one another; it is, in general, not possible t o compare or combine techniques. In this paper, we provide a framework to reason about many of the se these differing approaches using the unifying language of information theory.

In Data Science, evaluating model performance is very important and the most commonly used performance metric is the classification score. However, when dealing with fraud datasets with heavy class imbalance, a classification score does not make much sense. Instead, Receiver Operating Characteristic or ROC curves offer a better alternative. ROC is a plot of signal (True Positive Rate) against noise (False Positive Rate). The model performance is determined by looking at the area under the ROC curve (or AUC).

Supervised Learning is the crown jewel of Machine Learning. A couple years ago, Harvard Business Review released an article with the following title "Data Scientist: The Sexiest Job of the 21st Century." Ever since its release, Data Science or Statistics Departments become widely pursued by college students and, and Data Scientists (Nerds), for the first time, is referred to as being sexy. For some industries, Data Scientists have reshaped the corporation structure and reallocated a lot of decision-makings to the "front-line" workers. Being able to generate useful business insights from data has never been so easy.

Supervised Learning is the crown jewel of Machine Learning. A couple years ago, Harvard Business Review released an article with the following title "Data Scientist: The Sexiest Job of the 21st Century." Ever since its release, Data Science or Statistics Departments become widely pursued by college students and, and Data Scientists (Nerds), for the first time, is referred to as being sexy. For some industries, Data Scientists have reshaped the corporation structure and reallocated a lot of decision-makings to the "front-line" workers. Being able to generate useful business insights from data has never been so easy.

Uncertainty quantification for deep learning is a challenging open problem. Bayesian statistics offer a mathematically grounded framework to reason about uncertainties; however, approximate posteriors for modern neural networks still require prohibitive computational costs. We propose a family of algorithms which split the classification task into two stages: representation learning and uncertainty estimation. We compare four specific instances, where uncertainty estimation is performed via either an ensemble of Stochastic Gradient Descent or Stochastic Gradient Langevin Dynamics snapshots, an ensemble of bootstrapped logistic regressions, or via a number of Monte Carlo Dropout passes. We evaluate their performance in terms of \emph{selective} classification (risk-coverage), and their ability to detect out-of-distribution samples. Our experiments suggest there is limited value in adding multiple uncertainty layers to deep classifiers, and we observe that these simple methods strongly outperform a vanilla point-estimate SGD in some complex benchmarks like ImageNet.