Learning representation has been proven to be helpful in numerous machine learning tasks. The success of the majority of existing representation learning approaches often requires a large amount of consistent and noise-free labels. However, labels are not accessible in many real-world scenarios and they are usually annotated by the crowds. In practice, the crowdsourced labels are usually inconsistent among crowd workers given their diverse expertise and the number of crowdsourced labels is very limited. Thus, directly adopting crowdsourced labels for existing representation learning algorithms is inappropriate and suboptimal. In this paper, we investigate the above problem and propose a novel framework of \textbf{R}epresentation \textbf{L}earning with crowdsourced \textbf{L}abels, i.e., "RLL", which learns representation of data with crowdsourced labels by jointly and coherently solving the challenges introduced by limited and inconsistent labels. The proposed representation learning framework is evaluated in two real-world education applications. The experimental results demonstrate the benefits of our approach on learning representation from limited labeled data from the crowds, and show RLL is able to outperform state-of-the-art baselines. Moreover, detailed experiments are conducted on RLL to fully understand its key components and the corresponding performance.

Traffic sign recognition is an important component of many advanced driving assistance systems, and it is required for full autonomous driving. Computational performance is usually the bottleneck in using large scale neural networks for this purpose. SqueezeNet is a good candidate for efficient image classification of traffic signs, but in our experiments it does not reach high accuracy, and we believe this is due to lack of data, requiring data augmentation. Generative adversarial networks can learn the high dimensional distribution of empirical data, allowing the generation of new data points. In this paper we apply pix2pix GANs architecture to generate new traffic sign images and evaluate the use of these images in data augmentation. We were motivated to use pix2pix to translate symbolic sign images to real ones due to the mode collapse in Conditional GANs. Through our experiments we found that data augmentation using GAN can increase classification accuracy for circular traffic signs from 92.1% to 94.0%, and for triangular traffic signs from 93.8% to 95.3%, producing an overall improvement of 2%. However some traditional augmentation techniques can outperform GAN data augmentation, for example contrast variation in circular traffic signs (95.5%) and displacement on triangular traffic signs (96.7 %). Our negative results shows that while GANs can be naively used for data augmentation, they are not always the best choice, depending on the problem and variability in the data.

Which area in NYC is the most similar to Lower East Side? What about the NoHo Arts District in Los Angeles? Traditionally this task utilizes information about the type of places located within the areas and some popularity/quality metric. We take a different approach. In particular, urban dwellers' time-variant mobility is a reflection of how they interact with their city over time. Hence, in this paper, we introduce an approach, namely hood2vec, to identify the similarity between urban areas through learning a node embedding of the mobility network captured through Foursquare check-ins. We compare the pairwise similarities obtained from hood2vec with the ones obtained from comparing the types of venues in the different areas. The low correlation between the two indicates that the mobility dynamics and the venue types potentially capture different aspects of similarity between urban areas.

At Ahold Delhaize, there is an interest in using more complex machine learning techniques for sales forecasting. It is difficult to convince analysts, along with their superiors, to adopt these techniques since the models are considered to be 'black boxes,' even if they perform better than current models in use. We aim to explore the impact of contrastive explanations about large errors on users' attitudes towards a 'black-box' model. In this work, we make two contributions. The first is an algorithm, Monte Carlo Bounds for Reasonable Predictions (MC-BRP). Given a large error, MC-BRP determines (1) feature values that would result in a reasonable prediction, and (2) general trends between each feature and the target, based on Monte Carlo simulations. The second contribution is the evaluation of MC-BRP along with its outcomes, which has both objective and subjective components. We evaluate on a real dataset with real users from Ahold Delhaize by conducting a user study to determine if explanations generated by MC-BRP help users understand why a prediction results in a large error, and if this promotes trust in an automatically-learned model. The study shows that users are able to answer objective questions about the model's predictions with overall 81.7% accuracy when provided with these contrastive explanations. We also show that users who saw MC-BRP explanations understand why the model makes large errors in predictions significantly more than users in the control group.

In modern recommender systems, both users and items are associated with rich side information, which can help understand users and items. Such information is typically heterogeneous and can be roughly categorized into flat and hierarchical side information. While side information has been proved to be valuable, the majority of existing systems have exploited either only flat side information or only hierarchical side information due to the challenges brought by the heterogeneity. In this paper, we investigate the problem of exploiting heterogeneous side information for recommendations. Specifically, we propose a novel framework jointly captures flat and hierarchical side information with mathematical coherence. We demonstrate the effectiveness of the proposed framework via extensive experiments on various real-world datasets. Empirical results show that our approach is able to lead a significant performance gain over the state-of-the-art methods.

We explore the task of local music recommendation: provide listeners with personalized playlists of relevant tracks by artists who play most of their live events within a small geographic area. Most local artists tend to be obscure, long-tail artists and generally have little or no available user preference data associated with them. This creates a cold-start problem for collaborative filtering-based recommendation algorithms that depend on large amounts of such information to make accurate recommendations. In this paper, we compare the performance of three standard recommender system algorithms (Item-Item Neighborhood (IIN), Alternating Least Squares for Implicit Feedback (ALS), and Bayesian Personalized Ranking (BPR)) on the task of local music recommendation using the Million Playlist Dataset. To do this, we modify the standard evaluation procedure such that the algorithms only rank tracks by local artists for each of the eight different cities. Despite the fact that techniques based on matrix factorization (ALS, BPR) typically perform best on large recommendation tasks, we find that the neighborhood-based approach (IIN) performs best for long-tail local music recommendation.

Among the machine learning applications to business, recommender systems would take one of the top places when it comes to success and adoption. They help the user in accelerating the process of search while helping businesses maximize sales. Post phenomenal success in computer vision and speech recognition, deep learning methods are beginning to get applied to recommender systems. Current survey papers on deep learning in recommender systems provide a historical overview and taxonomy of recommender systems based on type. Our paper addresses the gaps of providing a taxonomy of deep learning approaches to address recommender systems problems in the areas of cold start and candidate generation in recommender systems. We outline different challenges in recommender systems into those related to the recommendations themselves (include relevance, speed, accuracy and scalability), those related to the nature of the data (cold start problem, imbalance and sparsity) and candidate generation. We then provide a taxonomy of deep learning techniques to address these challenges. Deep learning techniques are mapped to the different challenges in recommender systems providing an overview of how deep learning techniques can be used to address them. We contribute a taxonomy of deep learning techniques to address the cold start and candidate generation problems in recommender systems. Cold Start is addressed through additional features (for audio, images, text) and by learning hidden user and item representations. Candidate generation has been addressed by separate networks, RNNs, autoencoders and hybrid methods. We also summarize the advantages and limitations of these techniques while outlining areas for future research.

The anomaly detection of time series is a hotspot of time series data mining. The own characteristics of different anomaly detectors determine the abnormal data that they are good at. There is no detector can be optimizing in all types of anomalies. Moreover, it still has difficulties in industrial production due to problems such as a single detector can't be optimized at different time windows of the same time series. This paper proposes an adaptive model based on time series characteristics and selecting appropriate detector and run-time parameters for anomaly detection, which is called ATSDLN(Adaptive Time Series Detector Learning Network). We take the time series as the input of the model, and learn the time series representation through FCN. In order to realize the adaptive selection of detectors and run-time parameters according to the input time series, the outputs of FCN are the inputs of two sub-networks: the detector selection network and the run-time parameters selection network. In addition, the way that the variable layer width design of the parameter selection sub-network and the introduction of transfer learning make the model be with more expandability. Through experiments, it is found that ATSDLN can select appropriate anomaly detector and run-time parameters, and have strong expandability, which can quickly transfer. We investigate the performance of ATSDLN in public data sets, our methods outperform other methods in most cases with higher effect and better adaptation. We also show experimental results on public data sets to demonstrate how model structure and transfer learning affect the effectiveness.

Transformer is a popularly used neural network architecture, especially for language understanding. We introduce an extended and unified architecture which can be used for tasks involving a variety of modalities like image, text, videos, etc. We propose a spatio-temporal cache mechanism that enables learning spatial dimension of the input in addition to the hidden states corresponding to the temporal input sequence. The proposed architecture further enables a single model to support tasks with multiple input modalities as well as asynchronous multi-task learning, thus we refer to it as OmniNet. For example, a single instance of OmniNet can concurrently learn to perform the tasks of part-of-speech tagging, image captioning, visual question answering and video activity recognition. We demonstrate that training these four tasks together results in about three times compressed model while retaining the performance in comparison to training them individually. We also show that using this neural network pre-trained on some modalities assists in learning an unseen task. This illustrates the generalization capacity of the self-attention mechanism on the spatio-temporal cache present in OmniNet.

Often domain adaptation is performed using a discriminator (domain One such adversarial domain-invariant feature learning method classifier) to learn domain-invariant feature representations is the domain-adversarial neural network (DANN) [14, 15], which so that a classifier trained on labeled source data will generalize is a typical baseline for other variants. This method consists of a feature well to unlabeled target data. A line of research stemming from extractor network followed by two additional networks: a task semi-supervised learning uses pseudo labeling to directly generate classifier and a domain classifier (Figure 1). The network is updated "pseudo labels" for the unlabeled target data and trains a classifier by two competing objectives: (1) the feature extractor followed by on the now-labeled target data, where the samples are selected the task classifier learns to correctly classify the labeled source data or weighted based on some measure of confidence. In this paper, while the domain classifier learns to correctly predict whether the we propose multi-purposing the discriminator to not only aid in features originated from source or target data, and (2) the feature producing domain-invariant representations but also to provide extractor learns to make the domain classifier predict the domain pseudo labeling confidence.