The headline story here is that for the first time a system has been developed that exceeds human performance in one of the most difficult of all human speech recognition tasks: natural conversations held over the telephone. This is known as conversational telephone speech, or CTS. The reference datasets for this task are the Switchboard and Fisher data collections from the 1990s and early 2000s. The apocryphal story here is that human performance on the task is about 4% error rate. But no-one can quite pin down where that 4% number comes from.

In a way, AI is about understanding, and then mimicking how we think, learn and process information. The science and applications of AI have evolved since the early years: 1950's 1980's 2010's Generation 1 (From 1950's): Rule Based Systems (No Learning) In the early days, most applications of AI were rule-based computer programs (commonly known as Expert Systems) designed to solve problems that human brains performed easily. Such AI programs required experts to develop rules and combine with programs to solve problems. It required a programmer to write a program to capture the knowledge of a subject matter expert. The program then asked a series of questions to a user (usually not an expert in that subject) and then based on the answers/input provided, the computer would suggest a "solution" to the problem.

The key to getting good at applied machine learning is practicing on lots of different datasets. This is because each problem is different, requiring subtly different data preparation and modeling methods. In this post, you will discover 10 top standard machine learning datasets that you can use for practice. Each dataset is summarized in a consistent way. This makes them easy to compare and navigate for you to practice a specific data preparation technique or modeling method. Each dataset is small enough to fit into memory and review in a spreadsheet.

Now you can use Chatbots to get news and information in a personalized pattern. Famous media companies like CNN and Fox news have already launched their Chatbots on platforms like Facebook Messenger, Line and Kik as well as on voice-activated devices like Amazon's Alexa. Facebook has unveiled new capabilities for businesses and publishers on Messenger, enabling users to chat with CNN to get breaking news and personalized stories. People will now be able to chat with the companies and publishers like they would do with their friends. CNN is using chatterbots for Facebook Messenger to interact with users in a natural and human-like way.

The high growth rate of the hardware market for deep learning is attributed to the growing need for hardware platforms with a high computing power to run deep learning algorithms. There is increasing competition among established as well as startup players, leading to new product developments including both hardware development and software platforms to run deep learning algorithms and programs. For instance, Graphcore (a U.K.-based company) is developing the intelligent processing unit (IPU) for machine learning technology for use in applications from driverless cars to cloud computing. Some of the companies involved in the development of hardware for the deep learning technique are Google, Inc. (U.S.), Microsoft Corporation (U.S.), Intel Corporation (U.S.), Qualcomm, Inc. (U.S.), IBM Corporation (U.S.), and others.

Lately, machine learning, deep learning and other concepts which relate to data being used in order to adapt itself automatically have been very popular. Although this is pretty cool to be able to make an algorithm adapt itself automatically and this can be practical in some situations when humans do not have the capability to created the statistical models to use in order to analyse the available data, this is a mistake to think that this will replace every other data analysis techniques anytime soon because of some critical points. First, what are Artificial Intelligence, Machine Learning and Deep Learning? I guess that I have already scared people less familiar with data science techniques. Artificial intelligence (AI) is a set of rules that can be applied to a large set of data, it can be very advanced such as an autonomous car, which could decide to break or accelerate based on real time feedback from thousand of variables.

Chinese search giant Baidu Inc. is publicly releasing Chinese language APIs for its primary speech technologies as it and rivals continue to open up their portfolios of AI-based deep learning technologies. The Beijing-based company (NASDAQ: BIDU) said the APIs cover two variations of speech recognition, a speech synthesis tool and a fourth called "wake word" used to activate voice-controlled digital assistants such as the Amazon Echo. The speech APIs represent the latest in a series of public releases by the Chinese company that includes natural language processing along with facial and optical character recognition technologies. It also underscores how AI technology pioneers are attempting to attract application developers as competition heats up in the personal digital assistant and other emerging markets. In September, Baudu released a deep learning framework called "PaddlePaddle" that it pitches as a simple-to-use platform for developing new products and services.

Machine learning may once have been a topic of discussion only for computer scientists and researchers. Now, however, it is a technology businesses are eager to use. The need for machine learning and Artificial Intelligence (AI) is being driven by the massive amount of data being generated today. Statisticians can get insight from this data. But the volume is so large and growing at such a rate, the best way to tackle it is using the very same machines that are in part responsible for creating the data.

TechCrunch Disrupt in London is on December 5-6. As well as speakers and panels, we'll be featuring some demos by some amazing tech companies. The first will be by Boston Dynamics. Yes, folks, delegates to Disrupt London will get to see one of those amazing BD robots up close and personal, almost literally in the flesh (if they had any flesh, that is). Marc Raibert, CTO and Founder of Boston Dynamics will be demonstrating one of the amazing robots his team has created, but you'll have to come to find out which one… Raibert was Professor of Electrical Engineering and Computer Science at MIT and a member of the Artificial Intelligence Laboratory from 1986 through 1995.

Google's AI is not just better at grasping languages like Mandarin, but can now translate between two languages it hasn't even trained on. In a research paper, Google reveals how it uses its own "interlingua" to internally represent phrases, regardless of the language. The resulting "zero-shot" deep learning lets it translate a language pair with "reasonable" accuracy, as long as it has translated them both into another common language. The company recently switched its Translate feature to the deep-learning Google Neural Machine Translation (GNMT) system. That's an "end-to-end learning framework that learns from millions of examples," the company says, and has drastically improved translation quality.