These notes accompany the Stanford CS class CS231n: Convolutional Neural Networks for Visual Recognition. You can also submit a pull request directly to our git repo. We encourage the use of the hypothes.is

Microsoft has teamed up with the Tate's IK Prize winner for 2016, Italian-based communication research centre Fabrica, to surface images from Tate's extensive collection of more than 30,000 images that relate to other photos, supplied by Reuters, of current events as part of an art installation at the Tate called Recognition. The search engine uses Microsoft's Cognitive Services AI to recognize expressions, themes, context and style similarities between pictures in the news and Tate's massive collection, merging the modern and historical. The algorithm is not only able to match pictures, but also to explain why it thinks they are a matches, for example having similar objects, colours or compositions, and builds on technology such as HowOldAmI.net The exhibition is a show-case of Microsoft's AI technology and is also designed to improve public perception of Artificial Intelligence. "From a Microsoft perspective AI is the future – we're betting the company on this technology that extends the reach of humanity," said Dave Coplin, chief envisaging officer at the tech brand.

While chatbots have been around since the early 90's, this year they became completely synonymous with Facebook Messenger. Facebook brought automated, chat-based customer service into the mainstream at a time when businesses and buyers are obsessed with improving the customer experience. This technology will be transformative. It's part of a wave of innovations that is going to help businesses meet the expectations of their consumers. Who wouldn't want every interaction with a business to be faster, easier, and more like talking to a friend?

Researchers at Stanford University are utilizing artificial intelligence (AI) to identify areas of poverty in hard-to-reach places. Publishing their research in Science Magazine, the team consisting of Neal Jean, Marshall Burke, Michael Xie, Matthew Davis, David Lobell, and Stefano Ermon used deep learning algorithms to sort through millions of satellite images to identify economic conditions in five African countries. This research supports a forecast Tractica made over a year ago in our Artificial Intelligence for Enterprise Applications report, that spending on AI software by philanthropy organizations will grow dramatically over the next 10 years. Traditionally, philanthropy organizations have conducted door-to-door surveys to identify people living in poverty, but these surveys are imprecise, time-consuming, and expensive. Many international aid organizations including the World Bank have been trying to use satellite surveys to gather data remotely on developing countries, but the expense of gathering and analyzing this data using conventional methods has proven prohibitive.

Advancements in artificial intelligence, coupled with the proliferation of messaging apps, are fueling the development of chatbots -- software programs that use messaging as the interface through which to carry out any number of tasks, from scheduling a meeting, to reporting weather, to helping users buy a pair of shoes. Foreseeing immense potential, businesses are starting to invest heavily in the burgeoning bot economy. A number of brands and publishers have already deployed bots on messaging and collaboration channels, including HP, 1-800-Flowers, and CNN. While the bot revolution is still in the early phase, many believe 2016 will be the year these conversational interactions take off. In a new report from BI Intelligence, we explore the growing and disruptive bot landscape by investigating what bots are, how businesses are leveraging them, and where they will have the biggest impact.

Data clustering is the process of grouping items together based on similarities between the items of a group. Clustering can be used for data compression, data mining, pattern recognition, and machine learning. Examples of applications include clustering consumers into market segments, classifying manufactured units by their failure signatures, identifying crime hot spots, and identifying regions with similar geographical characteristics. Once clusters are defined, the next step may be to build a predictive model. TIBCO Spotfire makes it easy to perform clustering with these two popular out of box user-friendly solutions: 1. K-means Clustering 2. Hierarchical Clustering The k-means method is a popular and simple approach to perform clustering and Spotfire line charts help visualize data before performing calculations.

Lenovo this week unveiled a new addition to the Moto Z family at IFA 2016. It's the Moto Z Play, and as you might expect from similarly-named phones from recent history, this model is made to be more affordable than the flagships it shares its name with. Get a closer look in our hands-on video.

The Weka machine learning workbench is an easy to use and powerful platform for applied machine learning. Even though it is easy to use, you may still require some help or advice when using it on your own problems. In this post you will discover resources that you can use to get more help with Weka. How to Get More Help For the Weka Machine Learning Workbench Photo by Susanne Nilsson, some rights reserved. Your installation of Weka provides documentation that you can consult for simpler questions around how to use the platform.

A couple weeks ago,we discussed the concepts of both linear classification and parameterized learning. This type of learning allows us to take a set of input data and class labels, and actually learn a function that maps the input to the output predictions, simply by defining a set of parameters and optimizing over them. Our linear classification tutorial focused mainly on the concept of a scoring function and how it can be used to map the input data to class labels. But in order to actually "learn" the mapping from the input data to class labels, we need to discuss two important concepts: In today and next week's blog posts, we'll be discussing two loss functions that you'll commonly see when applying Machine Learning, Neural Networks, and Deep Learning algorithms: To learn more about your first loss function, Multi-class SVM loss, just keep reading. At the most basic level, a loss function is simply used to quantify how "good" or "bad" a given predictor is at classifying the input data points in a dataset.

The huge amount of data we generate is a mixed blessing. It gives us better insight into human behavior, but it's also a lot more stuff to sort through. For helping us wade through our data, and make better use of it, we'll turn to AI (artificial intelligence) for things as complicated as navigating the sales journey to just figuring out what the doggone temperature is outside. He talks about how our relationship with our data will change as we develop better tools to interact with it, and better insights to learn from it. Coenraets: "I've been building apps for many, many years using all the different technologies that came and went and looking at the different ways people access information," Coenraets says.