We're going to begin to integrate big data, artificial intelligence and machine learning into the provision and production of our goods and services. How we treat our ownership of structured and unstructured data.

Some say people are born leaders. Some say they're groomed to be leaders. Now, artificial intelligence is helping to coach managers into leaders, by teaching them soft skills like clear communication, teamwork, and problem solving. According to the Harvard Business Review, most companies wait 10 years to train managers. If you combine this with the fact that a top reason people leave jobs is due to poor managers, it's clear why building positive soft skills within managers is so important.

It began as all well-meaning ideas do -- as an effort to make life a little bit easier. This year I began building my own stock-picking artificial intelligence (AI) program. I created it to help me invest my savings and gave it a name: AlphaBean. But as AlphaBean became smarter and smarter, I eventually began to wonder: Could it replace me? My job involves studying companies and writing about them for the wider public. If AlphaBean could learn to invest, might it, or something like it, steal my job? Of course, I'm not the first person to create an investing algorithm. Algorithmic thinking about investing has existed for a long time -- much longer, in fact, than people normally think.

It's December - you know what that means. Everyone wants to talk about the new year. I guess you can include me in "everyone", because my main goal this December is to get you and your marketing strategy ready for 2018. We're going to kick it off by talking about marketing trends we expect to see in the coming year. Some of these are general marketing predictions for 2018, while others have a specific local focus; still, they will all probably affect your business in some way.

I know how Beam Search work, and i know that at each step of decoder, we keep k top result and continue decode with them. The thing i want to ask is beam search is applied to the test time only or in both test and train????????

Both of the major smart home platforms have a long-running problem with "discoverability": it's very hard to let users know what their devices can do, particularly if they're always improving thanks to rapid software updates. Amazon and Google are constantly experimenting with ways to get around this, but so far they have been timid. Amazon sends a weekly email, while Google includes some tips in its app. Expect to see them be bolder, particularly as powerful rivals such as Apple appear on the scene with worse AI but better sound. So don't be surprised if your Google Home or Amazon Echo begin to talk back, rather than simply following commands.

My pick for this year: "The Shattered Gradients Problem: If resnets are the answer, then what is the question?" Honorable mentions: 1. Poincaré Embeddings for Learning Hierarchical Representations (for elegance) 2. Inferring and Executing Programs for Visual Reasoning (for trying to tackle an important problem (not just VQA itself) the hard, but ultimately right way) 3. Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methods (for bringing empirical joy to my heart)

ZDNet's Sandra Vogel posted a formal review of the Huawei Mate 10 Pro, giving it an outstanding 9/10 rating. I've been spending quality time with this business-ready powerhouse and think that the AI found in the camera is worth discussing in a bit more detail. Huawei's partnership with Leica has resulted in some fantastic cameras and performance that is tough to beat. DxOMark awarded the Mate 10 Pro it second highest overall score, 97, and best still image score, 100. Keep in mind, these scores are not scaled to 100.

Results aren't very good, but I wonder if that's a result in and of itself? One thing that's been annoying me recently is that with all the advancements in the field in the past few years many people have started to think of research only being about finding the next AlexNet that will make waves throughout the field. I feel like research should be approached in a scientific manner with the goal of experimenting with the application of models and that all results are good results. Knowing that a specific approach doesn't work allows for us to find different methods without having to repeat the work other's have already done. I'd say these results definitely could be worth publishing, but I think you should do more work and testing before you consider that.

Label distribution learning (LDL) is a general learning framework, which assigns to an instance a distribution over a set of labels rather than a single label or multiple labels. Current LDL methods have either restricted assumptions on the expression form of the label distribution or limitations in representation learning, e.g., to learn deep features in an end-to-end manner. This paper presents label distribution learning forests (LDLFs) - a novel label distribution learning algorithm based on differentiable decision trees, which have several advantages: 1) Decision trees have the potential to model any general form of label distributions by a mixture of leaf node predictions. 2) The learning of differentiable decision trees can be combined with representation learning. We define a distribution-based loss function for a forest, enabling all the trees to be learned jointly, and show that an update function for leaf node predictions, which guarantees a strict decrease of the loss function, can be derived by variational bounding. The effectiveness of the proposed LDLFs is verified on several LDL tasks and a computer vision application, showing significant improvements to the state-of-the-art LDL methods.