arXiv subjects

Ran El-Yaniv

Publications and source records attributed to Ran El-Yaniv.

49 records · Page 3Linked to original sources

arXiv2017

Deep Learning of Robotic Tasks without a Simulator using Strong and Weak Human Supervision

We propose a scheme for training a computerized agent to perform complex human tasks such as highway steering. The scheme is designed to follow a natural learning process whereby a human instructor teaches a computerized trainee. The learning process consists of five elements: (i) unsupervised feature learning; (ii) supervised imitation learning; (iii) supervised reward induction; (iv) supervised safety module construction; and (v) reinforcement learning. We implemented the last four elements of the scheme using deep convolutional networks and applied it to successfully create a computerized agent capable of autonomous highway steering over the well-known racing game Assetto Corsa. We demonstrate that the use of the last four elements is essential to effectively carry out the steering task using vision alone, without access to a driving simulator internals, and operating in wall-clock time. This is made possible also through the introduction of a safety network, a novel way for preventing the agent from performing catastrophic mistakes during the reinforcement learning stage.

cs.AI↗

arXiv2017

Multi-Objective Non-parametric Sequential Prediction

Online-learning research has mainly been focusing on minimizing one objective function. In many real-world applications, however, several objective functions have to be considered simultaneously. Recently, an algorithm for dealing with several objective functions in the i.i.d. case has been presented. In this paper, we extend the multi-objective framework to the case of stationary and ergodic processes, thus allowing dependencies among observations. We first identify an asymptomatic lower bound for any prediction strategy and then present an algorithm whose predictions achieve the optimal solution while fulfilling any continuous and convex constraining criterion.

cs.LG↗

arXiv2016

Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations

We introduce a method to train Quantized Neural Networks (QNNs) --- neural networks with extremely low precision (e.g., 1-bit) weights and activations, at run-time. At train-time the quantized weights and activations are used for computing the parameter gradients. During the forward pass, QNNs drastically reduce memory size and accesses, and replace most arithmetic operations with bit-wise operations. As a result, power consumption is expected to be drastically reduced. We trained QNNs over the MNIST, CIFAR-10, SVHN and ImageNet datasets. The resulting QNNs achieve prediction accuracy comparable to their 32-bit counterparts. For example, our quantized version of AlexNet with 1-bit weights and 2-bit activations achieves $51\%$ top-1 accuracy. Moreover, we quantize the parameter gradients to 6-bits as well which enables gradients computation using only bit-wise operation. Quantized recurrent neural networks were tested over the Penn Treebank dataset, and achieved comparable accuracy as their 32-bit counterparts using only 4-bits. Last but not least, we programmed a binary matrix multiplication GPU kernel with which it is possible to run our MNIST QNN 7 times faster than with an unoptimized GPU kernel, without suffering any loss in classification accuracy. The QNN code is available online.

cs.NE↗

arXiv2016

Online Learning of Commission Avoidant Portfolio Ensembles

We present a novel online ensemble learning strategy for portfolio selection. The new strategy controls and exploits any set of commission-oblivious portfolio selection algorithms. The strategy handles transaction costs using a novel commission avoidance mechanism. We prove a logarithmic regret bound for our strategy with respect to optimal mixtures of the base algorithms. Numerical examples validate the viability of our method and show significant improvement over the state-of-the-art.

cs.AI↗

arXiv2016

Online Learning of Portfolio Ensembles with Sector Exposure Regularization

We consider online learning of ensembles of portfolio selection algorithms and aim to regularize risk by encouraging diversification with respect to a predefined risk-driven grouping of stocks. Our procedure uses online convex optimization to control capital allocation to underlying investment algorithms while encouraging non-sparsity over the given grouping. We prove a logarithmic regret for this procedure with respect to the best-in-hindsight ensemble. We applied the procedure with known mean-reversion portfolio selection algorithms using the standard GICS industry sector grouping. Empirical Experimental results showed an impressive percentage increase of risk-adjusted return (Sharpe ratio).

cs.LG↗

arXiv2016

Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1

We introduce a method to train Binarized Neural Networks (BNNs) - neural networks with binary weights and activations at run-time. At training-time the binary weights and activations are used for computing the parameters gradients. During the forward pass, BNNs drastically reduce memory size and accesses, and replace most arithmetic operations with bit-wise operations, which is expected to substantially improve power-efficiency. To validate the effectiveness of BNNs we conduct two sets of experiments on the Torch7 and Theano frameworks. On both, BNNs achieved nearly state-of-the-art results over the MNIST, CIFAR-10 and SVHN datasets. Last but not least, we wrote a binary matrix multiplication GPU kernel with which it is possible to run our MNIST BNN 7 times faster than with an unoptimized GPU kernel, without suffering any loss in classification accuracy. The code for training and running our BNNs is available on-line.

cs.LG↗

arXiv2015

Learn on Source, Refine on Target:A Model Transfer Learning Framework with Random Forests

We propose novel model transfer-learning methods that refine a decision forest model M learned within a "source" domain using a training set sampled from a "target" domain, assumed to be a variation of the source. We present two random forest transfer algorithms. The first algorithm searches greedily for locally optimal modifications of each tree structure by trying to locally expand or reduce the tree around individual nodes. The second algorithm does not modify structure, but only the parameter (thresholds) associated with decision nodes. We also propose to combine both methods by considering an ensemble that contains the union of the two forests. The proposed methods exhibit impressive experimental results over a range of problems.

cs.LG↗

arXiv2014

A Compression Technique for Analyzing Disagreement-Based Active Learning

We introduce a new and improved characterization of the label complexity of disagreement-based active learning, in which the leading quantity is the version space compression set size. This quantity is defined as the size of the smallest subset of the training data that induces the same version space. We show various applications of the new characterization, including a tight analysis of CAL and refined label complexity bounds for linear separators under mixtures of Gaussians and axis-aligned rectangles under product densities. The version space compression set size, as well as the new characterization of the label complexity, can be naturally extended to agnostic learning problems, for which we show new speedup results for two well known active learning algorithms.

cs.LG↗

arXiv2014

Transductive Rademacher Complexity and its Applications

We develop a technique for deriving data-dependent error bounds for transductive learning algorithms based on transductive Rademacher complexity. Our technique is based on a novel general error bound for transduction in terms of transductive Rademacher complexity, together with a novel bounding technique for Rademacher averages for particular algorithms, in terms of their "unlabeled-labeled" representation. This technique is relevant to many advanced graph-based transductive algorithms and we demonstrate its effectiveness by deriving error bounds to three well known algorithms. Finally, we present a new PAC-Bayesian bound for mixtures of transductive algorithms based on our Rademacher bounds.

cs.LG↗

arXiv2013

Semantic Sort: A Supervised Approach to Personalized Semantic Relatedness

We propose and study a novel supervised approach to learning statistical semantic relatedness models from subjectively annotated training examples. The proposed semantic model consists of parameterized co-occurrence statistics associated with textual units of a large background knowledge corpus. We present an efficient algorithm for learning such semantic models from a training sample of relatedness preferences. Our method is corpus independent and can essentially rely on any sufficiently large (unstructured) collection of coherent texts. Moreover, the approach facilitates the fitting of semantic models for specific users or groups of users. We present the results of extensive range of experiments from small to large scale, indicating that the proposed method is effective and competitive with the state-of-the-art.

cs.CL↗

arXiv2012

Autoregressive short-term prediction of turning points using support vector regression

This work is concerned with autoregressive prediction of turning points in financial price sequences. Such turning points are critical local extrema points along a series, which mark the start of new swings. Predicting the future time of such turning points or even their early or late identification slightly before or after the fact has useful applications in economics and finance. Building on recently proposed neural network model for turning point prediction, we propose and study a new autoregressive model for predicting turning points of small swings. Our method relies on a known turning point indicator, a Fourier enriched representation of price histories, and support vector regression. We empirically examine the performance of the proposed method over a long history of the Dow Jones Industrial average. Our study shows that the proposed method is superior to the previous neural network model, in terms of trading performance of a simple trading application and also exhibits a quantifiable advantage over the buy-and-hold benchmark.

cs.LG↗

arXiv2010

On the Foundations of Adversarial Single-Class Classification

Motivated by authentication, intrusion and spam detection applications we consider single-class classification (SCC) as a two-person game between the learner and an adversary. In this game the learner has a sample from a target distribution and the goal is to construct a classifier capable of distinguishing observations from the target distribution from observations emitted from an unknown other distribution. The ideal SCC classifier must guarantee a given tolerance for the false-positive error (false alarm rate) while minimizing the false negative error (intruder pass rate). Viewing SCC as a two-person zero-sum game we identify both deterministic and randomized optimal classification strategies for different game variants. We demonstrate that randomized classification can provide a significant advantage. In the deterministic setting we show how to reduce SCC to two-class classification where in the two-class problem the other class is a synthetically generated distribution. We provide an efficient and practical algorithm for constructing and solving the two class problem. The algorithm distinguishes low density regions of the target distribution and is shown to be consistent.

cs.LG↗

arXiv2010

Hierarchical Multiclass Decompositions with Application to Authorship Determination

This paper is mainly concerned with the question of how to decompose multiclass classification problems into binary subproblems. We extend known Jensen-Shannon bounds on the Bayes risk of binary problems to hierarchical multiclass problems and use these bounds to develop a heuristic procedure for constructing hierarchical multiclass decomposition for multinomials. We test our method and compare it to the well known "all-pairs" decomposition. Our tests are performed using a new authorship determination benchmark test of machine learning authors. The new method consistently outperforms the all-pairs decomposition when the number of classes is small and breaks even on larger multiclass problems. Using both methods, the classification accuracy we achieve, using an SVM over a feature set consisting of both high frequency single tokens and high frequency token-pairs, appears to be exceptionally high compared to known results in authorship determination.

cs.AI↗