Yes, to be fair a lot of the points are valid, but they come from a wishy-washy, non-technical angle (one could argue this is because of his intended audience, but I submit that part of this is because of his lack of ML knowledge). And they are already well-known.

I've got this ominous feeling that 2018 could be the year that everything just changes dramatically. The incredible breakthroughs we saw in 2017 for Deep Learning is going to carry over in a very powerful way in 2018. A lot of work coming from research will be migrating itself into everyday software applications. As I've done last year, here are my predictions for 2018. Many Deep Learning hardware startup ventures will begin to finally deliver their silicon in 2018.

As a machine learning engineer, I have been learning and playing with deep learning for quite some time. Now, after finishing all Andrew NG newest deep learning courses in Coursera, I decided to put some of my understanding of this field into a blog post. I found writing things down is an efficient way in subduing a topic. In addition, I hope that this post might be useful to those who want to get started into Deep Learning. Alright, so let us talk about deep learning.

Malware recognition modules decide if an object is a threat, based on the data they have collected on it. This data may be collected at different phases: – Pre-execution phase data is anything you can tell about a file without executing it. This may include executable file format descriptions, code descriptions, binary data statistics, text strings and information extracted via code emulation and other similar data. In the early epochs of the cyber era, the number of malware threats was relatively low, and simple handcrafted pre-execution rules were often enough to detect threats. But a decade ago, the tremendous growth of the malware stream did not allow anti-malware solutions to rely solely on the expensive manual creation of detection rules. It was natural for anti-malware companies to start augmenting their malware detection and classification with machine learning, a computer science area that has shown great success in image recognition, searching and decision- making. Machine Learning Methods for Malware Detection In this article, we summarize our decade's worth of experience with implementing machine learning into protecting our customers from cyberthreats. In other words, a machine learning algorithm discovers and formalizes the principles that underlie the data it sees. With this knowledge, the algorithm can reason the properties of previously unseen samples. In malware detection, a previously unseen sample could be a new file. Its hidden property could be malware or benign. A mathematically formalized set of principles underlying data properties is called the model. Machine learning has a broad variety of approaches that it takes to a solution rather than a single method. These approaches have different capacities and different tasks that they suit best. Unsupervised learning One machine learning approach is unsupervised learning. In this setting, we are given only a data set without the right answers for the task. The goal is to discover the structure of the data or the law of data generation. One important example is clustering. Clustering is a task that includes splitting a data set into groups of similar objects. Another task is representation learning – this includes building an informative feature set for objects based on their low- level description (for example, an autoencoder model). Large unlabeled datasets are available to cybersecurity vendors and the cost of their manual labeling by experts is high – this makes unsupervised learning valuable for threat detection. Clustering can help with optimizing efforts for the manual labeling of new samples. With informative embedding, we can decrease the number of labeled objects needed for the usage of the next machine learning approach in our pipeline: supervised learning.

The process of transforming observed data into predictive mathematical models of the physical world has always been paramount in science and engineering. Although data is currently being collected at an ever-increasing pace, devising meaningful models out of such observations in an automated fashion still remains an open problem. In this work, we put forth a machine learning approach for identifying nonlinear dynamical systems from data. Specifically, we blend classical tools from numerical analysis, namely the multi-step time-stepping schemes, with powerful nonlinear function approximators, namely deep neural networks, to distill the mechanisms that govern the evolution of a given data-set. We test the effectiveness of our approach for several benchmark problems involving the identification of complex, nonlinear and chaotic dynamics, and we demonstrate how this allows us to accurately learn the dynamics, forecast future states, and identify basins of attraction. In particular, we study the Lorenz system, the fluid flow behind a cylinder, the Hopf bifurcation, and the Glycoltic oscillator model as an example of complicated nonlinear dynamics typical of biological systems.

Abstract-- Recently, Neural networks have seen a huge surge in its adoption due to their ability to provide high accuracy on various tasks. On the other hand, the existence of adversarial examples have raised suspicions regarding the generalization capabilities of neural networks. In this work, we focus on the weight matrix learnt by the neural networks and hypothesize that ill conditioned weight matrix is one of the contributing factors in neural network's susceptibility towards adversarial examples. For ensuring that the learnt weight matrix's condition number remains sufficiently low, we suggest using orthogonal regularizer. We show that this indeed helps in increasing the adversarial accuracy on MNIST and F-MNIST datasets.

Interior tomography for the region-of-interest (ROI) imaging has advantages of using a small detector and reducing X-ray radiation dose. However, standard analytic reconstruction suffers from severe cupping artifacts due to existence of null space in the truncated Radon transform. Existing penalized reconstruction methods may address this problem but they require extensive computations due to the iterative reconstruction. Inspired by the recent deep learning approaches to low-dose and sparse view CT, here we propose a deep learning architecture that removes null space signals from the FBP reconstruction. Experimental results have shown that the proposed method provides near-perfect reconstruction with about 7-10 dB improvement in PSNR over existing methods in spite of significantly reduced run-time complexity.

Structure based ligand discovery is one of the most successful approaches for augmenting the drug discovery process. Currently, there is a notable shift towards machine learning (ML) methodologies to aid such procedures. Deep learning has recently gained considerable attention as it allows the model to "learn" to extract features that are relevant for the task at hand. We have developed a novel deep neural network estimating the binding affinity of ligand-receptor complexes. The complex is represented with a 3D grid, and the model utilizes a 3D convolution to produce a feature map of this representation, treating the atoms of both proteins and ligands in the same manner. Our network was tested on the CASF "scoring power" benchmark and Astex Diverse Set and outperformed classical scoring functions.

Cortical circuits exhibit intricate recurrent architectures that are remarkably similar across different brain areas. Such stereotyped structure suggests the existence of common computational principles. However, such principles have remained largely elusive. Inspired by gated-memory networks, namely long short-term memory networks (LSTMs), we introduce a recurrent neural network in which information is gated through inhibitory cells that are subtractive (subLSTM). We propose a natural mapping of subLSTMs onto known canonical excitatory-inhibitory cortical microcircuits. Our empirical evaluation across sequential image classification and language modelling tasks shows that subLSTM units can achieve similar performance to LSTM units. These results suggest that cortical circuits can be optimised to solve complex contextual problems and proposes a novel view on their computational function. Overall our work provides a step towards unifying recurrent networks as used in machine learning with their biological counterparts.

We describe a deep learning approach for automated brain hemorrhage detection from computed tomography (CT) scans. Our model emulates the procedure followed by radiologists to analyse a 3D CT scan in real-world. Similar to radiologists, the model sifts through 2D cross-sectional slices while paying close attention to potential hemorrhagic regions. Further, the model utilizes 3D context from neighboring slices to improve predictions at each slice and subsequently, aggregates the slice-level predictions to provide diagnosis at CT level. We refer to our proposed approach as Recurrent Attention DenseNet (RADnet) as it employs original DenseNet architecture along with adding the components of attention for slice level predictions and recurrent neural network layer for incorporating 3D context. The real-world performance of RADnet has been benchmarked against independent analysis performed by three senior radiologists for 77 brain CTs. RADnet demonstrates 81.82% hemorrhage prediction accuracy at CT level that is comparable to radiologists. Further, RADnet achieves higher recall than two of the three radiologists, which is remarkable.