There's a lot of chatter about chatbots these days and how we might be able to use them in the future. The biggest question seems to be whether chatbots can be useful enough to convincingly replace human conversation. Before chatbots can reach that point, they'll need to develop and mature into a technology that enables human communication with a computer using natural language. Most bots today are not at the level where they can flawlessly replicate conversation. Some chatbots today are not fed enough data.

Sony is looking at taking the Xperia brand beyond smartphones and is showing off some prototype gadgets at the IFA trade show in Berlin. It was keen to underline that not all of them might make it to manufacturing, but those on display were fully working prototypes with hardware and software stable enough that the company let a reporter test them out. This is an Android-based computer with a short-throw projector in a single box. The projector sends the screen image onto a nearby tabletop or wall and users can interact with it by using their fingers on the projected image. An infrared sensor in the device watches the fingers and figures out where they are pressing on the projected screen.

Not content with helping cure cancers and winning Jeopardy, Watson wants to get inside our heads and our homes, whispering instructions into our wireless headsets and helping us do our laundry. IBM will work with appliance maker Whirlpool, TV and camera company Panasonic, wireless headphone designer Bragi and Withings owner Nokia to add Watson's cognitive computing capabilities to their products, the company said. Those cognitive capabilities could help devices talk with one another, or with us. For example, a washing machine could tell a dryer what program to use for the clothes it has just washed, or tell its owner when to order more detergent. Computer vision techniques could help security cameras distinguish between friends and strangers or identify suspicious activity.

A universal unanswered question in neuroscience and machine learning is whether computers can decode the patterns of the human brain. Multi-Voxels Pattern Analysis (MVPA) is a critical tool for addressing this question. However, there are two challenges in the previous MVPA methods, which include decreasing sparsity and noises in the extracted features and increasing the performance of prediction. In overcoming mentioned challenges, this paper proposes Anatomical Pattern Analysis (APA) for decoding visual stimuli in the human brain. This framework develops a novel anatomical feature extraction method and a new imbalance AdaBoost algorithm for binary classification. Further, it utilizes an Error-Correcting Output Codes (ECOC) method for multi-class prediction. APA can automatically detect active regions for each category of the visual stimuli. Moreover, it enables us to combine homogeneous datasets for applying advanced classification. Experimental studies on 4 visual categories (words, consonants, objects and scrambled photos) demonstrate that the proposed approach achieves superior performance to state-of-the-art methods.

Have you ever looked at a machine learning classification model and thought, I could have made that? Well, that is what we test in this project, comparing XGBoost trained on human engineered features to training directly on data. The human engineered features do not outperform XGBoost trained di- rectly on the data, but they are comparable. This project con- tributes a novel method for utilizing human created classifi- cation models on high dimensional datasets.

We present an information-theoretic framework for solving global black-box optimization problems that also have black-box constraints. Of particular interest to us is to efficiently solve problems with decoupled constraints, in which subsets of the objective and constraint functions may be evaluated independently. For example, when the objective is evaluated on a CPU and the constraints are evaluated independently on a GPU. These problems require an acquisition function that can be separated into the contributions of the individual function evaluations. We develop one such acquisition function and call it Predictive Entropy Search with Constraints (PESC). PESC is an approximation to the expected information gain criterion and it compares favorably to alternative approaches based on improvement in several synthetic and real-world problems. In addition to this, we consider problems with a mix of functions that are fast and slow to evaluate. These problems require balancing the amount of time spent in the meta-computation of PESC and in the actual evaluation of the target objective. We take a bounded rationality approach and develop partial update for PESC which trades off accuracy against speed. We then propose a method for adaptively switching between the partial and full updates for PESC. This allows us to interpolate between versions of PESC that are efficient in terms of function evaluations and those that are efficient in terms of wall-clock time. Overall, we demonstrate that PESC is an effective algorithm that provides a promising direction towards a unified solution for constrained Bayesian optimization.

We provide two fundamental results on the population (infinite-sample) likelihood function of Gaussian mixture models with $M \geq 3$ components. Our first main result shows that the population likelihood function has bad local maxima even in the special case of equally-weighted mixtures of well-separated and spherical Gaussians. We prove that the log-likelihood value of these bad local maxima can be arbitrarily worse than that of any global optimum, thereby resolving an open question of Srebro (2007). Our second main result shows that the EM algorithm (or a first-order variant of it) with random initialization will converge to bad critical points with probability at least $1-e^{-\Omega(M)}$. We further establish that a first-order variant of EM will not converge to strict saddle points almost surely, indicating that the poor performance of the first-order method can be attributed to the existence of bad local maxima rather than bad saddle points. Overall, our results highlight the necessity of careful initialization when using the EM algorithm in practice, even when applied in highly favorable settings.

Today, a lot of high-profile deep machine learning projects in the cloud are powered by GPUs; specifically NVIDIA GPUs. Even Facebook uses them for its own machine learning work behind-the-scenes. GPUs are able to handle the massive amounts of computing power required to train deep neural networks that facilitate these projects. But Huawei deputy chairman and rotating CEO Eric Xu believes the future of machine learning lies in dedicated processors. Read on to find out more. In the past few years, NVIDIA has made a big push to dethrone CPUs in the deep machine learning space.

Apple haters have always made the case that the company's massive success is as much the product of marketing and salesmanship as it is any kind of technical innovation. Whatever else Apple cofounder Steve Jobs was, he was the consummate salesman. Maybe the original iPhone could have sold itself back in 2007, but Jobs' legendary introductory event definitely helped. But the world has changed. As smartphone innovation seems to have plateaued, the tech giants of the world, notably Google, Microsoft, and Facebook, have doubled down on machine learning and artificial intelligence -- the trendy technology that's making for smarter, more personalized apps and devices.

In this special guest feature, Victor Amin, Data Scientist at SendGrid, advises that businesses implementing machine learning systems focus on data quality first and worry about algorithms later in order to ensure accuracy and reliability in production. After graduating cum laude from Princeton University, Victor earned a PhD studying applications of machine learning to quantum chemistry at Northwestern University. At SendGrid, Victor builds machine learning models to predict engagement and detect abuse in a mailstream that handles over a billion emails per day. It's obvious that you need data before you can implement a machine learning system, but project planners often overlook questions regarding training set collection, cleaning, and maintenance. There are so many sources of big data in today's business systems that it seems like getting enough of the right data ought to be easy!