Welcome to part five of Learning AI if You Suck at Math. If you missed parts 1, 2, 3, 4, 6, and 7 be sure to check them out! Convnets are the workhorses of computer vision. At TensorFlow Summit 2017, a researcher showed how they're using a convnet to detect skin cancer as well as a dermatologist with a smart phone! So why are neural networks so powerful? So what's pattern recognition and why do we care if it's automatic? Patterns come in many forms but let's take two critical examples: In image processing pattern recognition is known as feature extraction. When you look at a photo or something in the real world you're selectively picking out the key features that allow you to make sense of it. This is something you do unconsciously. When you see the picture of my cat Dove you think "cat" or "awwwwww" but you don't really know how you do that. Your built in neural network knows this is a cat. It seems simple to you because you do it every day, but that's because the complexity is hidden away from you. Your brain is a black box. You come with no instruction manual.

Throughout history, there have been moments when the progress of technology has taken great steps forward, when a combination of the right tools, a capacity for innovation, and sparks of ingenuity lead to breakthroughs that transform how we live our lives. How we produce and process information is critical to innovation – and our methods of recording and communicating information have themselves undergone great leaps. From the development of writing, to Gutenberg's printing press – which advanced the spread of knowledge to the masses and ushered in the enlightenment and scientific revolution – to the first programmable digital computer Colossus, the cost of reproducing and communicating information, or data, has fallen again and again. At the same time, tools for processing and making sense of large quantities of data have developed exponentially – with artificial intelligence (AI) representing the latest leap. In the same way that Gutenberg's press ushered in a new era of growth, data-driven technologies such as AI will underpin our future prosperity. There is no doubt that machine learning and AI is already improving peoples' lives, from intelligent personal assistants that can prepare us for changes in the weather, to systems that protect our money from criminals, or devices that offer medical advice from the comfort of our own home. And this is only the start; the potential of AI is undeniable. Our next challenge will be to harness this technology to transform how we diagnose diseases, manufacture goods and build our homes. Using advanced algorithmic techniques such as'deep learning', AI has the potential to solve complex problems fast, and in so doing, free up time and raise productivity. But we also need to make sure AI benefits everyone in the UK, which is why – in addition to this Sector Deal – the government is establishing a Centre for Data Ethics and Innovation to advise on the ethical use of data, including for AI. The huge global opportunity AI presents is why the Industrial Strategy white paper identified AI and data as 1 of 4 Grand Challenges – in which the UK can lead the world for years to come.

In order to build deep neural networks, one modification to the basic convolutional operation that we have to use is padding. Let's see how it works. What we saw in earlier posts is that if we take a \(6 \times 6 \) image and convolve it with a \(3 \times 3 \) filter, we end up with a \(4 \times 4 \) output (or with a \(4 \times 4 \) matrix), and that's because the number of possible positions for our \(3 \times 3 \) filter, to fit in our \(6 \times 6 \) matrix is \(4 \times 4 \) . If we convolve an \(n\times n\) image with an \(f\times f\) filter, what are the dimensions of the output matrix? If we have an \(n\times n\), and convolve that with a \(f\times f \), then the dimension of the output will be: \((n-f 1)\times (n-f 1) \) .

This post is in construction. Subscribe to our paid newsletter to see the latest trend and tutorials in Machine Learning, Deep Learning. Our best, high quality contents, notes from conferences are for paid subscribers only. There is a free subscribe tier as well. Subscribers, please remember to check the website for the latest articles. We avoid sending all articles to your inbox so not spam you.

Ed Note: Gideon is speaking at ODSC Europe 2019, see his talk "Augmented Programming" there. Over the past decade, deep learning research has led to significant advances in perceptual tasks, such as object detection, face recognition, and speech recognition. In each of these use cases, raw real-world inputs have to be mapped into a normalized representation. Deep learning has also started to lead to significant advances in natural language processing, for example, where contextual embeddings enable multi-task transfer learning. A newer, emerging field for machine learning is in its application within the production, deployment, and maintenance of software.

Computer chipmaker and storage company Micron Technology Inc. is pitching its hardware for artificial intelligence workloads after acquiring a startup called FWDNXT. The company announced the acquisition at its annual Micron Insight conference in San Francisco today, describing FWDNXT as a provider of AI hardware and software for deep learning, which is a subset of AI that tries to mimic the way the human brain solves problems. Micron's plan is to integrate FWDNXT's technology with its own, optimized flash memory products to create what it says will be a "comprehensive AI development platform." "FWDNXT is an architecture designed to create fast-time-to-market edge AI solutions through an extremely easy to use software framework with broad modeling support and flexibility," Micron Executive Vice President and Chief Business Officer Sumit Sadana said in a statement. "FWDNXT's five generations of machine learning inference engine development and neural network algorithms, combined with Micron's deep memory expertise, unlocks new power and performance capabilities to enable innovation for the most complex and demanding edge applications."

Real-world datasets are often biased with respect to key demographic factors such as race and gender. Due to the latent nature of the underlying factors, detecting and mitigating bias is especially challenging for unsupervised machine learning. We present a weakly supervised algorithm for overcoming dataset bias for deep generative models. Our approach requires access to an additional small, unlabeled but unbiased dataset as the supervision signal, thus sidestepping the need for explicit labels on the underlying bias factors. Using this supplementary dataset, we detect the bias in existing datasets via a density ratio technique and learn generative models which efficiently achieve the twin goals of: 1) data efficiency by using training examples from both biased and unbiased datasets for learning, 2) unbiased data generation at test time. Empirically, we demonstrate the efficacy of our approach which reduces bias w.r.t. latent factors by 57.1% on average over baselines for comparable image generation using generative adversarial networks.

Recent years have witnessed growing interest in machine learning-based models and techniques for low-dose X-ray CT (LDCT) imaging tasks. The methods can typically be categorized into supervised learning methods and unsupervised or model-based learning methods. Supervised learning methods have recently shown success in image restoration tasks. However, they often rely on large training sets. Model-based learning methods such as dictionary or transform learning do not require large or paired training sets and often have good generalization properties, since they learn general properties of CT image sets. Recent works have shown the promising reconstruction performance of methods such as PWLS-ULTRA that rely on clustering the underlying (reconstructed) image patches into a learned union of transforms. In this paper, we propose a new Supervised-UnsuPERvised (SUPER) reconstruction framework for LDCT image reconstruction that combines the benefits of supervised learning methods and (unsupervised) transform learning-based methods such as PWLS-ULTRA that involve highly image-adaptive clustering. The SUPER model consists of several layers, each of which includes a deep network learned in a supervised manner and an unsupervised iterative method that involves image-adaptive components. The SUPER reconstruction algorithms are learned in a greedy manner from training data. The proposed SUPER learning methods dramatically outperform both the constituent supervised learning-based networks and iterative algorithms for LDCT, and use much fewer iterations in the iterative reconstruction modules.

--Federated learning (FL) has recently emerged as a new form of collaborative machine learning, where a common model can be learned while keeping all the training data on local devices. Although it is designed for enhancing the data privacy, we demonstrated in this paper a new direction in inference attacks in the context of FL, where valuable information about training data can be obtained by adversaries with very limited power . In particular, we proposed three new types of attacks to exploit this vulnerability. The first type of attack, Class Sniffing, can detect whether a certain label appears in training. The other two types of attacks can determine the quantity of each label, i.e., Quantity Inference attack determines the composition proportion of the training label owned by the selected clients in a single round, while Whole Determination attack determines that of the whole training process. We evaluated our attacks on a variety of tasks and datasets with different settings, and the corresponding results showed that our attacks work well generally. Finally, we analyzed the impact of major hyper-parameters to our attacks and discussed possible defenses. The emergence of federated learning (FL) enables multiple devices to learn a common model while keeping all the training data on their own devices. It allows for less resource consumption on the cloud and ensures the privacy at the same time. Multiple applications have benefited from FL, including mobile phones [1, 2, 3], wearable devices [4, 5], autonomous vehicles [6, 7], etc. In standard federated learning, all participants are required to train their local models. A random subset of clients will be selected each round, who will upload their gradient updates to the central server. Similar FL architectures can be found in [8, 9, 10, 11, 12, 13]. One interesting question here is about the security and privacy implication in the FL training process. Any characteristic of clients' private data needs to be protected carefully since it may reveal some important private information about the training data - e.g., the distribution of labels might show the diversity of participants. Similarly, what the training data consists of is also what attackers want to explore, i.e., can they determine the quantity proportion of different labels in the whole training dataset during the training process?