Online multi-kernel learning is promising in the era of mobile computing, in which a combined classifier with multiple kernels are offline trained, and online adapts to personalized features for serving the end user precisely and smartly. The online adaptation is mainly carried out at the end-devices, which requires the adaptation algorithms to be light, efficient and accurate. Previous results focused mainly on efficiency. This paper proposes an novel online model adaptation framework for not only efficiency but also optimal online adaptation. At first, an online optimal incremental simplex tableau (IST)algorithm is proposed, which approaches the model adaption by linear programming and produces the optimized model update in each step when a personalized training data is collected.But keeping online optimal in each step is expensive and may cause over-fitting especially when the online data is noisy. A Fast-IST approach is therefore proposed, which measures the deviation between the training data and the current model. It schedules updating only when enough deviation is detected. The efficiency of each update is further enhanced by running IST only limited iterations, which bounds the computation complexity. Theoretical analysis and extensive evaluations show that Fast-IST saves computation cost greatly, while achieving speedy and accurate model adaptation.It provides better model adaptation speed and accuracy while using even lower computing cost than the state-of-the art.

A major challenge in the training of recurrent neural networks is the so-called vanishing or exploding gradient problem. The use of a norm-preserving transition operator can address this issue, but parametrization is challenging. In this work we focus on unitary operators and describe a parametrization using the Lie algebra u( n ) associated with the Lie group U ( n ) of n × n unitary matrices. The exponential map provides a correspondence between these spaces, and allows us to define a unitary matrix using n 2 real coefficients relative to a basis of the Lie algebra. The parametrization is closed under additive updates of these coefficients, and thus provides a simple space in which to do gradient descent. We demonstrate the effectiveness of this parametrization on the problem of learning arbitrary unitary operators, comparing to several baselines and outperforming a recently-proposed lower-dimensional parametrization. We additionally use our parametrization to generalize a recently-proposed unitary recurrent neural network to arbitrary unitary matrices, using it to solve standard long-memory tasks.

Online joint parameter and state estimation is a core problem for temporal models.Most existing methods are either restricted to a particular class of models (e.g., the Storvik filter) or computationally expensive (e.g., particle MCMC). We propose a novel nearly-black-box algorithm, the Assumed Parameter Filter (APF), a hybrid of particle filtering for state variables and assumed density filtering for parameter variables.It has the following advantages:(a) it is online and computationally efficient;(b) it is applicable to both discrete and continuous parameter spaces with arbitrary transition dynamics.On a variety of toy and real models, APF generates more accurate results within a fixed computation budget compared to several standard algorithms from the literature.

This paper is about the estimation of the maximum expected value of an infinite set of random variables.This estimation problem is relevant in many fields, like the Reinforcement Learning (RL) one.In RL it is well known that, in some stochastic environments, a bias in the estimation error can increase step-by-step the approximation error leading to large overestimates of the true action values. Recently, some approaches have been proposed to reduce such bias in order to get better action-value estimates, but are limited to finite problems.In this paper, we leverage on the recently proposed weighted estimator and on Gaussian process regression to derive a new method that is able to natively handle infinitely many random variables.We show how these techniques can be used to face both continuous state and continuous actions RL problems.To evaluate the effectiveness of the proposed approach we perform empirical comparisons with related approaches.

In this paper, we consider the problem of actively learning a linear classifier through query synthesis where the learner can construct artificial queries in order to estimate the true decision boundaries. This problem has recently gained a lot of interest in automated science and adversarial reverse engineering for which only heuristic algorithms are known. In such applications, queries can be constructed de novo to elicit information (e.g., automated science) or to evade detection with minimal cost (e.g., adversarial reverse engineering). We develop a general framework, called dimension coupling (DC), that 1) reduces a d-dimensional learning problem to d-1 low dimensional sub-problems, 2) solves each sub-problem efficiently, 3) appropriately aggregates the results and outputs a linear classifier, and 4) provides a theoretical guarantee for all possible schemes of aggregation. The proposed method is proved resilient to noise. We show that the DC framework avoids the curse of dimensionality: its computational complexity scales linearly with the dimension. Moreover, we show that the query complexity of DC is near optimal (within a constant factor of the optimum algorithm). To further support our theoretical analysis, we compare the performance of DC with the existing work. We observe that DC consistently outperforms the prior arts in terms of query complexity while often running orders of magnitude faster.

The problem of multi-instance multi-label learning (MIML) requires a bag of instances to be assigned a set of labels most relevant to the bag as a whole. The problem finds numerous applications in machine learning, computer vision, and natural language processing settings where only partial or distant supervision is available. We present a novel method for optimizing multivariate performance measures in the MIML setting. Our approach MIML-perf uses a novel plug-in technique and offers a seamless way to optimize a vast variety of performance measures such as macro and micro-F measure, average precision, which are performance measures of choice in multi-label learning domains. MIML-perf offers two key benefits over the state of the art. Firstly, across a diverse range of benchmark tasks, ranging from relation extraction to text categorization and scene classification, MIML-perf offers superior performance as compared to state of the art methods designed specifically for these tasks. Secondly, MIML-perf operates with significantly reduced running times as compared to other methods, often by an order of magnitude or more.

Accurately predicting students' future performance based on their tracked academic records in college programs is crucial for effectively carrying out necessary pedagogical interventions to ensure students' on-time graduation. Although there is a rich literature on predicting student performance in solving problems and studying courses using data-driven approaches, predicting student performance in completing college programs is much less studied and faces new challenges, mainly due to the diversity of courses selected by students and the requirement of continuous tracking and incorporation of students' evolving progresses. In this paper, we develop a novel algorithm that enables progressive prediction of students' performance by adapting ensemble learning techniques and utilizing education-specific domain knowledge. We prove its prediction performance guarantee and show its performance improvement against benchmark algorithms on a real-world student dataset from UCLA.

With the expansion of data, increasing imbalanced data has emerged. When the imbalance ratio of data is high, most existing imbalanced learning methods decline in classification performance. To address this problem, a few highly imbalanced learning methods have been presented. However, most of them are still sensitive to the high imbalance ratio. This work aims to provide an effective solution for the highly imbalanced data classification problem. We conduct highly imbalanced learning from the perspective of feature learning. We partition the majority class into multiple blocks with each being balanced to the minority class and combine each block with the minority class to construct a balanced sample set. Multiset feature learning (MFL) is performed on these sets to learn discriminant features. We thus propose an uncorrelated cost-sensitive multiset learning (UCML) approach. UCML provides a multiple sets construction strategy, incorporates the cost-sensitive factor into MFL, and designs a weighted uncorrelated constraint to remove the correlation among multiset features. Experiments on five highly imbalanced datasets indicate that: UCML outperforms state-of-the-art imbalanced learning methods.

We propose a lifelong learning system that has the ability to reuse and transfer knowledge from one task to another while efficiently retaining the previously learned knowledge-base. Knowledge is transferred by learning reusable skills to solve tasks in Minecraft, a popular video game which is an unsolved and high-dimensional lifelong learning problem. These reusable skills, which we refer to as Deep Skill Networks, are then incorporated into our novel Hierarchical Deep Reinforcement Learning Network (H-DRLN) architecture using two techniques: (1) a deep skill array and (2) skill distillation, our novel variation of policy distillation (Rusu et. al. 2015) for learning skills. Skill distillation enables the H-DRLN to efficiently retain knowledge and therefore scale in lifelong learning, by accumulating knowledge and encapsulating multiple reusable skills into a single distilled network. The H-DRLN exhibits superior performance and lower learning sample complexity compared to the regular Deep Q Network (Mnih et. al. 2015) in sub-domains of Minecraft.

Automatic generation of natural language description for individual images (a.k.a. image captioning) has attracted extensive research attention. In this paper, we take one step further to investigate the generation of a paragraph to describe a photo stream for the purpose of storytelling. This task is even more challenging than individual image description due to the difficulty in modeling the large visual variance in an ordered photo collection and in preserving the long-term language coherence among multiple sentences. To deal with these challenges, we formulate the task as a sequence-to-sequence learning problem and propose a novel joint learning model by leveraging the semantic coherence in a photo stream. Specifically, to reduce visual variance, we learn a semantic space by jointly embedding each photo with its corresponding contextual sentence, so that the semantically related photos and their correlations are discovered. Then, to preserve language coherence in the paragraph, we learn a novel Bidirectional Attention-based Recurrent Neural Network (BARNN) model, which can attend on the discovered semantic relation to produce a sentence sequence and maintain its consistence with the photo stream. We integrate the two-step learning components into one single optimization formulation and train the network in an end-to-end manner. Experiments on three widely-used datasets (NYC/Disney/SIND) show that the proposed approach outperforms state-of-the-art methods with large margins for both retrieval and paragraph generation tasks. We also show the subjective preference of the machine-generated stories by the proposed approach over the baselines through a user study with 40 human subjects.