ML algorithms work only with numerical values. So there is a need to model a problem and its data completely in numbers. For example, to run a clustering algorithm on a road network, representing the network / graph as an adjacency matrix is one way to model it. Similarly, a tabular data with a mix of numerical and non-numerical / categorical data also needs to be transformed or encoded to a table of only numbers for a ML algorithm to work on. Columns of string values are quite common in tabular data and in this article, some ideas on how to encode them, especially ones with high cardinality and are of known lengths like IP addresses, mobile numbers etc. are discussed.

Affective computing [31] is "computing that relates to, arises from, or influences emotions." It is very important in human-machine interaction, as humans cannot have longlasting intimate relationships with machines if they cannot understand our affects and respond appropriately. Both affect classification and regression have been extensively studied in the literature [24], [43], [45], [46], [48]. For affect classification, the most commonly used categories are the six basic emotions (anger, disgust, fear, happiness, sadness, and surprise) proposed by Ekman et al. [5]. For regression, affects are usually represented as numbers in the 2D space of arousal and valence [35], or in the 3D space of arousal, valence, and dominance [25]. Recently, Yannakakis et al. [50] also argued that the nature of emotions is ordinal, and hence preference learning [51] should also play an important role in affective computing. Various input signals could be used in affective computing, e.g., speech [21], [47], facial expressions [8], [29], physiological signals [7], [43], and multimodal combination [26], [53]. Numerous features could be extracted from each modality. For example, 6,373 acoustic features were extracted by OpenSMILE [6] in the InterSpeech 2013 Computational Paralinguistics Challenge.

The multi-armed bandit is a sequential allocation task where an agent must learn a policy that maximizes long term payoff, where only the reward of the played arm is observed at each iteration. In the stochastic setting, the reward for each action is generated from an unknown distribution, which depends on a given 'context', available at each interaction with the world. Thompson sampling is a generative, interpretable multi-armed bandit algorithm that has been shown both to perform well in practice, and to enjoy optimality properties for certain reward functions. Nevertheless, Thompson sampling requires sampling from parameter posteriors and calculation of expected rewards, which are possible for a very limited choice of distributions. We here extend Thompson sampling to more complex scenarios by adopting a very flexible set of reward distributions: nonparametric Gaussian mixture models. The generative process of Bayesian nonparametric mixtures naturally aligns with the Bayesian modeling of multi-armed bandits. This allows for the implementation of an efficient and flexible Thompson sampling algorithm: the nonparametric model autonomously determines its complexity in an online fashion, as it observes new rewards for the played arms. We show how the proposed method sequentially learns the nonparametric mixture model that best approximates the true underlying reward distribution. Our contribution is valuable for practical scenarios, as it avoids stringent model specifications, and yet attains reduced regret.

Regression problems are pervasive in real-world applications. Generally a substantial amount of labeled samples are needed to build a regression model with good generalization ability. However, many times it is relatively easy to collect a large number of unlabeled samples, but time-consuming or expensive to label them. Active learning for regression (ALR) is a methodology to reduce the number of labeled samples, by selecting the most beneficial ones to label, instead of random selection. This paper proposes two new ALR approaches based on greedy sampling (GS). The first approach (GSy) selects new samples to increase the diversity in the output space, and the second (iGS) selects new samples to increase the diversity in both input and output spaces. Extensive experiments on 12 UCI and CMU StatLib datasets from various domains, and on 15 subjects on EEG-based driver drowsiness estimation, verified their effectiveness and robustness.

This paper presents a light-weight and accurate deep neural model for audiovisual emotion recognition. To design this model, the authors followed a philosophy of simplicity, drastically limiting the number of parameters to learn from the target datasets, always choosing the simplest earning methods: i) transfer learning and low-dimensional space embedding allows to reduce the dimensionality of the representations. ii) The isual temporal information is handled by a simple score-per-frame selection process, averaged across time. iii) A simple frame selection echanism is also proposed to weight the images of a sequence. iv) The fusion of the different modalities is performed at prediction level (late usion). We also highlight the inherent challenges of the AFEW dataset and the difficulty of model selection with as few as 383 validation equences. The proposed real-time emotion classifier achieved a state-of-the-art accuracy of 60.64 % on the test set of AFEW, and ranked 4th at he Emotion in the Wild 2018 challenge.

Acquisition of labeled training samples for affective computing is usually costly and time-consuming, as affects are intrinsically subjective, subtle and uncertain, and hence multiple human assessors are needed to evaluate each affective sample. Particularly, for affect estimation in the 3D space of valence, arousal and dominance, each assessor has to perform the evaluations in three dimensions, which makes the labeling problem even more challenging. Many sophisticated machine learning approaches have been proposed to reduce the data labeling requirement in various other domains, but so far few have considered affective computing. This paper proposes two multi-task active learning for regression approaches, which select the most beneficial samples to label, by considering the three affect primitives simultaneously. Experimental results on the VAM corpus demonstrated that our optimal sample selection approaches can result in better estimation performance than random selection and several traditional single-task active learning approaches. Thus, they can help alleviate the data labeling problem in affective computing, i.e., better estimation performance can be obtained from fewer labeling queries.

Almost all EEG-based brain-computer interfaces (BCIs) need some labeled subject-specific data to calibrate a new subject, as neural responses are different across subjects to even the same stimulus. So, a major challenge in developing high-performance and user-friendly BCIs is to cope with such individual differences so that the calibration can be reduced or even completely eliminated. This paper focuses on the latter. More specifically, we consider an offline application scenario, in which we have unlabeled EEG trials from a new subject, and would like to accurately label them by leveraging auxiliary labeled EEG trials from other subjects in the same task. To accommodate the individual differences, we propose a novel unsupervised approach to align the EEG trials from different subjects in the Euclidean space to make them more consistent. It has three desirable properties: 1) the aligned trial lie in the Euclidean space, which can be used by any Euclidean space signal processing and machine learning approach; 2) it can be computed very efficiently; and, 3) it does not need any labeled trials from the new subject. Experiments on motor imagery and event-related potentials demonstrated the effectiveness and efficiency of our approach.

Abstract--In near future, vulnerable road users (VRUs) such as cyclists and pedestrians will be equipped with smart devices and wearables which are capable to communicate with intelligent vehicles and other traffic participants. Road users are then able to cooperate on different levels, such as in cooperative intention detection for advanced VRU protection. Smart devices can be used to detect intentions, e.g., an occluded cyclist intending to cross the road, to warn vehicles of VRUs, and prevent potential collisions. This article presents a human activity recognition approach to detect the starting movement of cyclists wearing smart devices. We propose a novel two-stage feature selection procedure using a score specialized for robust starting detection reducing the false positive detections and leading to understandable and interpretable features. The detection is modelled as a classification problem and realized by means of a machine learning classifier. We introduce an auxiliary class, that models starting movements and allows to integrate early movement indicators, i.e., body part movements indicating future behaviour. In this way we improve the robustness and reduce the detection time of the classifier. Our empirical studies with real-world data originating from experiments which involve 49 test subjects and consists of 84 starting motions show that we are able to detect the starting movements early. Investigations concerning the device wearing location show that for devices worn in the trouser pocket the detector has less false detections and detects starting movements faster on average. We found that we can further improve the results when we train distinct classifiers for different wearing locations. In our work, we envision future mixed traffic scenarios where automated cars, trucks, sensor-equipped infrastructure, and other road users equipped with smart devices or other wearables are interconnected by means of ad hoc networks. This allows the traffic participants to cooperate, i.e., determine and maintain local models of the surrounding traffic situations. Vulnerable road users (VRUs) will still play an important role in future urban traffic.

Abstract--Distributed gradient descent (DGD) is an efficient way of implementing gradient descent (GD), especially for large data sets, by dividing the computation tasks into smaller sub-tasks and assigning to different computing servers (CSs) to be executed in parallel. In standard parallel execution, per-iteration waiting time is limited by the execution time of the straggling servers. Coded DGD techniques have been introduced recently, which can tolerate straggling servers via assigning redundant computation tasks to the CSs. In most of the existing DGD schemes, either with coded computation or coded communication, the non-straggling CSs transmit one message per iteration once they complete all their assigned computation tasks. However, although the straggling servers cannot complete all their assigned tasks, they are often able to complete a certain portion of them. In this paper, we allow multiple transmissions from each CS at each iteration in order to make sure a maximum number of completed computations can be reported to the aggregating server (AS), including the straggling servers. We numerically show that the average completion time per iteration can be reduced significantly by slightly increasing the communication load per server . Index Terms --Distributed gradient descent, coded computation, coded gradient, polynomial codes, maximum-distance separable codes.

Minimizing a convex, quadratic objective is a fundamental problem in machine learning and optimization. In this work, we study prove information-theoretic, gradient query complexity lower bounds for minimizing convex quadratic functions, which, unlike prior works, apply even for randomized algorithms. Specifically, we construct a distribution over quadratic functions that witnesses lower bounds which match those known for deterministic algorithms, up to multiplicative constants. The distribution which witnesses our lower bound is in fact quite benign: it is both closed form, and derived from classical ensembles in random matrix theory. We believe that our construction constitutes a plausible "average case" setting, and thus provides compelling evidence that the worst case and average case complexity of convex-quadratic optimization are essentially identical.