The paper "Model Compression" is available here: https://www.cs.cornell.edu/ We also thank Experiment for sponsoring our series. Subscribe if you would like to see more of these! - http://www.youtube.com/subscription_c... The thumbnail background image was created by John Lord - https://flic.kr/p/nVUaB

If I swap positive class and negative class, then train a model again (I tried decision tree, adaboost, svm from scikit-learn built-in package) for a two class classification problem. Sometimes, I can see AUC slightly change (around 1-2%). Anyone have any ideas why there are such changes? For ROC curve, x-axis is false positive rate, and y-axis and true positive rate. When prediction model gives prediction scores, we will order the scores from higher value to lower value, and then choose threshold according to the sorted values and calculate at the specific threshold point, what is the fpr and tpr.

Every organization becomes involved in the transition to the new cognitive computing technology epoch soon. This begs the question: "Does artificial intelligence need leadership?" We talked about future challenges in leadership and business with Udo Hornfeck, Vice President at LEONI, a tier-1-supplier in the automotive industry with approximately 70.000 employees worldwide and located in Kitzingen, Germany. Mr Hornfeck leads the Global Research & Development departments. Mark McGregor Leadership Center: Mr Hornfeck, how do you define leadership?

'Chatbots are the new apps," said Microsoft's CEO Satya Nadella earlier this year. He was not the first senior tech exec to make this claim. "Threads are the new apps," suggested Facebook's head of messaging products David Marcus in January, referring to the threads of conversation in apps such as Facebook Messenger and WhatsApp. Nadella and Marcus see chatbots – computer programs that you interact with by "chatting", for example in threads in messaging apps – as an important new human/machine interface. Both of their companies have launched tools to help developers create these bots, and between April and September, more than 30,000 were made for Facebook Messenger alone. The shopping and breaking news bots in Messenger's ancestors are chatbots such as AI psychotherapist Eliza from the mid-1960s and Parry, a bot mimicking a human with paranoid schizophrenia, in the early 1970s. Since 1991, the chatbot equivalent of the Olympics has been the annual Loebner prize, which challenges bots ...

A supervised machine learning model aims to learn a function f(x) y from a list of training pairs (x1,y1), (x2,y2), … for which data are recorded (Fig 1B). One typical application in biology is to predict the viability of a cancer cell line when exposed to a chosen drug (Menden et al, 2013; Eduati et al, 2015). The input features (x) would capture somatic sequence variants of the cell line, chemical make?up of the drug and its concentration, which together with the measured viability (output label y) can be used to train a support vector machine, a random forest classifier or a related method (functional relationship f). Given a new cell line (unlabelled data sample x*) in the future, the learnt function predicts its survival (output label y*) by calculating f(x*), even if f resembles more of a black box, and its inner workings of why particular mutation combinations influence cell growth are not easily interpreted. Both regression (where y is a real number) and classification (where y is a categorical class label) can be viewed in this way. As a counterpart, unsupervised machine learning approaches aim to discover patterns from the data samples x themselves, without the need for output labels y. Methods such as clustering, principal component analysis and outlier detection are typical examples of unsupervised models applied to biological data. The inputs x, calculated from the raw data, represent what the model "sees about the world", and their choice is highly problem?specific (Fig 1C). Deriving most informative features is essential for performance, but the process can be labour?intensive

Advances in machine learning have been driven by innovations and ideas from many fields. Inspired by the way that humans learn, Reinforcement Learning (RL) is concerned with algorithms which improve with trial-and-error feedback to optimize future performance. Board games and video games often have well-defined reward functions which allow for straightforward optimization with RL algorithms. Algorithmic advances have allowed for RL to be in real-world problems, such as high degree-of-freedom robotic manipulation and large-scale recommendation tasks, with more complex goals. Twitter Cortex invests in novel state-of-the-art machine learning methods to improve the quality of our products.

When Sebastian Thrun started working on self-driving cars at Google in 2007, few people outside of the company took him seriously. "I can tell you very senior CEOs of major American car companies would shake my hand and turn away because I wasn't worth talking to," said Thrun, now the co-founder and CEO of online higher education startup Udacity, in an interview with Recode earlier this week. A little less than a decade later, dozens of self-driving startups have cropped up while automakers around the world clamor, wallet in hand, to secure their place in the fast-moving world of fully automated transportation. And these companies are hungry for talent and skill sets many don't have. "Uber has just bought a half-a-year-old company [Otto] with 70 employees for almost 700 million," Thrun said.

After decades of stagnation in research and development of artificial intelligence solutions, machine learning has blossomed again – and its seeds are spreading to financial services. Sofia Lotto Persio reports on how AI is improving the field. Almost 20 years later, in March 2016, Google's AlphaGo programme beat South Korean champion Lee Sedol at four matches of Go, a strategy game, in what represents another historic breakthrough for artificial intelligence (AI). In the two-decade span between these achievements, AI has progressed tremendously. Go is a much more complex game than chess, as there are more possible positions on the board than there are atoms in the universe.

NOTE: Part 1 of 2: The following is an excerpt from Thomas International's white paper on Cognitive Fit. It can be found here: http://campaigns.thomasus.com/cognitive/ Employees with strong cognitive ability are a competitive advantage to an organization-- particularly when they are in roles that require quick learning, first-time problem solving, thinking "on your feet" and dealing with ambiguity. People who score well on cognitive ability tests are more likely to develop a greater knowledge of the job more quickly, make effective decisions and successfully reason and strategize to solve problems. In fact, a landmark study from 1998 reviewed 85 years of research and found that higher cognitive ability is linked with higher productivity and performance.

Bayes's Theorem fundamentally is based on the concept of "validity of Beliefs". Reverend Thomas Bayes was a Presbyterian minster and a Mathematician who pondered much about developing the proof of existence of God. He came up with the Theorem in 18th century (which was later refined by Pierre-Simmon Laplace) to fix or establish the validity of'existing' or'previous' Beliefs in the face of best available'new' evidence. Think of it as a equation to correct prior beliefs based on new evidence. One of the popular example used to explain Bayes's Theorem is to detect if a patient has a certain disease or not.