Food delivery company DoorDash says personalized restaurant recommendations based on AI are seeing a significant lift in orders, compared to regular recommendations based on popularity. In an interview with VentureBeat, DoorDash product manager Jimmy Liu said customers who saw personalized recommendations on average "were over 25 percent more likely" to place an order versus people who saw the most popular restaurants in their area. We talked with Liu on the eve of the company's announcement today that it's rolling out these machine-learning based recommendations to all of its users, after testing it on increasing percentages of its customer base. Millions of users have already seen the recommendations, the company said. Liu said the 25 percent lift from recommendations came specifically from email campaigns.

This paper introduces the kernel mixture network, a new method for nonparametric estimation of conditional probability densities using neural networks. We model arbitrarily complex conditional densities as linear combinations of a family of kernel functions centered at a subset of training points. The weights are determined by the outer layer of a deep neural network, trained by minimizing the negative log likelihood. This generalizes the popular quantized softmax approach, which can be seen as a kernel mixture network with square and non-overlapping kernels. We test the performance of our method on two important applications, namely Bayesian filtering and generative modeling. In the Bayesian filtering example, we show that the method can be used to filter complex nonlinear and non-Gaussian signals defined on manifolds. The resulting kernel mixture network filter outperforms both the quantized softmax filter and the extended Kalman filter in terms of model likelihood. Finally, our experiments on generative models show that, given the same architecture, the kernel mixture network leads to higher test set likelihood, less overfitting and more diversified and realistic generated samples than the quantized softmax approach.

Regulators require financial institutions to estimate counterparty default risks from liquid CDS quotes for the valuation and risk management of OTC derivatives. However, the vast majority of counterparties do not have liquid CDS quotes and need proxy CDS rates. Existing methods cannot account for counterparty-specific default risks; we propose to construct proxy CDS rates by associating to illiquid counterparty liquid CDS Proxy based on Machine Learning Techniques. After testing 156 classifiers from 8 most popular classifier families, we found that some classifiers achieve highly satisfactory accuracy rates. Furthermore, we have rank-ordered the performances and investigated performance variations amongst and within the 8 classifier families. This paper is, to the best of our knowledge, the first systematic study of CDS Proxy construction by Machine Learning techniques, and the first systematic classifier comparison study based entirely on financial market data. Its findings both confirm and contrast existing classifier performance literature. Given the typically highly correlated nature of financial data, we investigated the impact of correlation on classifier performance. The techniques used in this paper should be of interest for financial institutions seeking a CDS Proxy method, and can serve for proxy construction for other financial variables. Some directions for future research are indicated.

The idea of cross-validation is to split the data into N subsets, to put one subset aside, to estimate parameters of the model from the remaining N-1 subsets, and to use the retained subset to estimate the error of the model. Such a process is repeated N times - with each of the N subsets being used as the validation set . Then the values of the errors obtained in such N steps are combined to provide the final estimate of the model error. The cross-validation is used in various classification and prediction procedures, such as regression analysis, discriminant analysis, neural networks and classification and regression trees (CART) . The goal is to improve the quality of the decision that is made from the outcome of the study on the basis of statistical methods, and to ensure that maximum information is obtained from scarce experimental data.

The Monty Hall problem was first featured on the classic game show "Let's make a Deal". In the final segment of the show, contestants were presented with a choice of three different doors. Behind two of the doors would be a goat, and behind the third would be an extravagant prize such as a car. The contestant begins the game by picking one door. The host, Monty Hall, would then open one of the remaining doors.

Text mining (deriving information from text) is a wide field which has gained popularity with the huge text data being generated. Automation of a number of applications like sentiment analysis, document classification, topic classification, text summarization, machine translation, etc has been done using machine learning models. Spam filtering is a beginner's example of document classification task which involves classifying an email as spam or non-spam (a.k.a. Spam box in your Gmail account is the best example of this. So lets get started in building a spam filter on a publicly available mail corpus.

Machine Learning is a vast area of Computer Science that is concerned with designing algorithms which form good models of the world around us (the data coming from the world around us). Within Machine Learning many tasks are - or can be reformulated as - classification tasks. In classification tasks we are trying to produce a model which can give the correlation between the input data and the class each input belongs to. This model is formed with the feature-values of the input-data. For example, the dataset contains datapoints belonging to the classes Apples, Pears and Oranges and based on the features of the datapoints (weight, color, size etc) we are trying to predict the class. We need some amount of training data to train the Classifier, i.e. form a correct model of the data.

And why is it useful for BigData? 29 What is "continuous features binning"? What is a Standard Scaling? 38 Why are statistical distributions important? What is a Bias - Variance tradeoff? What is a training set, a validation set, a test set and a gold set in supervised and unsupervised learning? What is a cross-validation and what is an overfitting?

An important problem in Bayesian networks classifiers (BNC) is to discover relevant variables that can achieve optimal classification performance. We propose a method based on Bayesian inference for estimating and incorporating feature relevance in classification using BNCs. We empirically validate our method on an application to event recognition in natural images using object and scene information.

Interaction between humans and machines, like social robots, requires real time recognition of human actions. Most approaches to this problem wait for the end of the gesture to perform classification. In this paper we present a deep learning approach to online gesture recognition that allows for an estimation of the current gesture since its beginning. Our approach is to modify the existing Temporal Deep Belief Network (TDBN) architecture. The result is a Discriminative Temporal Deep Belief Network (DTDBN) which we apply to the online classification of motion capture streams. We optimize and evaluate our model in comparison with related work.