The key contribution of this paper is to propose a classification into two dimensions of the load forecasting studies to decide which forecasting tools to use in which case. This classification aims to provide a synthetic view of the relevant forecasting techniques and methodologies by forecasting problem. In addition, the key principles of the main techniques and methodologies used are summarized along with the reviews of these papers. The classification process relies on two couples of parameters that define a forecasting problem. Each article is classified with key information about the dataset used and the forecasting tools implemented: the forecasting techniques (probabilistic or deterministic) and methodologies, the data cleansing techniques, and the error metrics. The process to select the articles reviewed in this paper was conducted into two steps. First, a set of load forecasting studies was built based on relevant load forecasting reviews and forecasting competitions. The second step consisted in selecting the most relevant studies of this set based on the following criteria: the quality of the description of the forecasting techniques and methodologies implemented, the description of the results, and the contributions. This paper can be read in two passes. The first one by identifying the forecasting problem of interest to select the corresponding class into one of the four classification tables. Each one references all the articles classified across a forecasting horizon. They provide a synthetic view of the forecasting tools used by articles addressing similar forecasting problems. Then, a second level composed of four Tables summarizes key information about the forecasting tools and the results of these studies. The second pass consists in reading the key principles of the main techniques and methodologies of interest and the reviews of the articles.
Mental health poses a significant challenge for an individual's well-being. Text analysis of rich resources, like social media, can contribute to deeper understanding of illnesses and provide means for their early detection. We tackle a challenge of detecting social media users' mental status through deep learning-based models, moving away from traditional approaches to the task. In a binary classification task on predicting if a user suffers from one of nine different disorders, a hierarchical attention network outperforms previously set benchmarks for four of the disorders. Furthermore, we explore the limitations of our model and analyze phrases relevant for classification by inspecting the model's word-level attention weights.
Prediction models based on deep neural networks are increasingly gaining attention for fast and accurate virtual screening systems. For decision makings in virtual screening, researchers find it useful to interpret an output of classification system as probability, since such interpretation allows them to filter out more desirable compounds. However, probabilistic interpretation cannot be correct for models that hold over-parameterization problems or inappropriate regularizations, leading to unreliable prediction and decision making. In this regard, we concern the reliability of neural prediction models on molecular properties, especially when models are trained with sparse data points and imbalanced distributions. This work aims to propose guidelines for training reliable models, we thus provide methodological details and ablation studies on the following train principles. We investigate the effects of model architectures, regularization methods, and loss functions on the prediction performance and reliability of classification results. Moreover, we evaluate prediction reliability of models on virtual screening scenario. Our result highlights that correct choice of regularization and inference methods is evidently important to achieve high success rate, especially in data imbalanced situation. All experiments were performed under a single unified model implementation to alleviate external randomness in model training and to enable precise comparison of results.
Plant diseases serve as one of main threats to food security and crop production. It is thus valuable to exploit recent advances of artificial intelligence to assist plant disease diagnosis. One popular approach is to transform this problem as a leaf image classification task, which can be then addressed by the powerful convolutional neural networks (CNNs). However, the performance of CNN-based classification approach depends on a large amount of high-quality manually labeled training data, which are inevitably introduced noise on labels in practice, leading to model overfitting and performance degradation. To overcome this problem, we propose a novel framework that incorporates rectified meta-learning module into common CNN paradigm to train a noise-robust deep network without using extra supervision information. The proposed method enjoys the following merits: i) A rectified meta-learning is designed to pay more attention to unbiased samples, leading to accelerated convergence and improved classification accuracy. ii) Our method is free on assumption of label noise distribution, which works well on various kinds of noise. iii) Our method serves as a plug-and-play module, which can be embedded into any deep models optimized by gradient descent based method. Extensive experiments are conducted to demonstrate the superior performance of our algorithm over the state-of-the-arts.
This article introduces the solutions of the two champion teams, `MMfruit' for the detection track and `MMfruitSeg' for the segmentation track, in OpenImage Challenge 2019. It is commonly known that for an object detector, the shared feature at the end of the backbone is not appropriate for both classification and regression, which greatly limits the performance of both single stage detector and Faster RCNN \cite{ren2015faster} based detector. In this competition, we observe that even with a shared feature, different locations in one object has completely inconsistent performances for the two tasks. \textit{E.g. the features of salient locations are usually good for classification, while those around the object edge are good for regression.} Inspired by this, we propose the Decoupling Head (DH) to disentangle the object classification and regression via the self-learned optimal feature extraction, which leads to a great improvement. Furthermore, we adjust the soft-NMS algorithm to adj-NMS to obtain stable performance improvement. Finally, a well-designed ensemble strategy via voting the bounding box location and confidence is proposed. We will also introduce several training/inferencing strategies and a bag of tricks that give minor improvement. Given those masses of details, we train and aggregate 28 global models with various backbones, heads and 3+2 expert models, and achieves the 1st place on the OpenImage 2019 Object Detection Challenge on the both public and private leadboards. Given such good instance bounding box, we further design a simple instance-level semantic segmentation pipeline and achieve the 1st place on the segmentation challenge.
The ``shared head for classification and localization'' (sibling head), firstly denominated in Fast RCNN~\cite{girshick2015fast}, has been leading the fashion of the object detection community in the past five years. This paper provides the observation that the spatial misalignment between the two object functions in the sibling head can considerably hurt the training process, but this misalignment can be resolved by a very simple operator called task-aware spatial disentanglement (TSD). Considering the classification and regression, TSD decouples them from the spatial dimension by generating two disentangled proposals for them, which are estimated by the shared proposal. This is inspired by the natural insight that for one instance, the features in some salient area may have rich information for classification while these around the boundary may be good at bounding box regression. Surprisingly, this simple design can boost all backbones and models on both MS COCO and Google OpenImage consistently by ~3% mAP. Further, we propose a progressive constraint to enlarge the performance margin between the disentangled and the shared proposals, and gain ~1% more mAP. We show the \algname{} breaks through the upper bound of nowadays single-model detector by a large margin (mAP 49.4 with ResNet-101, 51.2 with SENet154), and is the core model of our 1st place solution on the Google OpenImage Challenge 2019.
Many text classification tasks are domain-dependent, and various domain adaptation approaches have been proposed to predict unlabeled data in a new domain. Domain-adversarial neural networks (DANN) and their variants have been used widely recently and have achieved promising results for this problem. However, most of these approaches assume that the label proportions of the source and target domains are similar, which rarely holds in most real-world scenarios. Sometimes the label shift can be large and the DANN fails to learn domain-invariant features. In this study, we focus on unsupervised domain adaptation of text classification with label shift and introduce a domain adversarial network with label proportions estimation (DAN-LPE) framework. The DAN-LPE simultaneously trains a domain adversarial net and processes label proportions estimation by the confusion of the source domain and the predictions of the target domain. Experiments show the DAN-LPE achieves a good estimate of the target label distributions and reduces the label shift to improve the classification performance.
Zero-shot learning strives to classify unseen categories for which no data is available during training. In the generalized variant, the test samples can further belong to seen or unseen categories. The state-of-the-art relies on Generative Adversarial Networks that synthesize unseen class features by leveraging class-specific semantic embeddings. During training, they generate semantically consistent features, but discard this constraint during feature synthesis and classification. We propose to enforce semantic consistency at all stages of (generalized) zero-shot learning: training, feature synthesis and classification. We first introduce a feedback loop, from a semantic embedding decoder, that iteratively refines the generated features during both the training and feature synthesis stages. The synthesized features together with their corresponding latent embeddings from the decoder are then transformed into discriminative features and utilized during classification to reduce ambiguities among categories. Experiments on (generalized) zero-shot object and action classification reveal the benefit of semantic consistency and iterative feedback, outperforming existing methods on six zero-shot learning benchmarks. Source code at https://github.com/akshitac8/tfvaegan.
The field of Few-Shot Learning (FSL), or learning from very few (typically $1$ or $5$) examples per novel class (unseen during training), has received a lot of attention and significant performance advances in the recent literature. While number of techniques have been proposed for FSL, several factors have emerged as most important for FSL performance, awarding SOTA even to the simplest of techniques. These are: the backbone architecture (bigger is better), type of pre-training on the base classes (meta-training vs regular multi-class, currently regular wins), quantity and diversity of the base classes set (the more the merrier, resulting in richer and better adaptive features), and the use of self-supervised tasks during pre-training (serving as a proxy for increasing the diversity of the base set). In this paper we propose yet another simple technique that is important for the few shot learning performance - a search for a compact feature sub-space that is discriminative for a given few-shot test task. We show that the Task-Adaptive Feature Sub-Space Learning (TAFSSL) can significantly boost the performance in FSL scenarios when some additional unlabeled data accompanies the novel few-shot task, be it either the set of unlabeled queries (transductive FSL) or some additional set of unlabeled data samples (semi-supervised FSL). Specifically, we show that on the challenging miniImageNet and tieredImageNet benchmarks, TAFSSL can improve the current state-of-the-art in both transductive and semi-supervised FSL settings by more than $5\%$, while increasing the benefit of using unlabeled data in FSL to above $10\%$ performance gain.
MRI (Magnetic Resonance Imaging) is a technique used to analyze and diagnose the problem defined by images like cancer or tumor in a brain. Physicians require good contrast images for better treatment purpose as it contains maximum information of the disease. MRI images are low contrast images which make diagnoses difficult; hence better localization of image pixels is required. Histogram Equalization techniques help to enhance the image so that it gives an improved visual quality and a well defined problem. The contrast and brightness is enhanced in such a way that it does not lose its original information and the brightness is preserved. We compare the different equalization techniques in this paper; the techniques are critically studied and elaborated. They are also tabulated to compare various parameters present in the image. In addition we have also segmented and extracted the tumor part out of the brain using K-means algorithm. For classification and feature extraction the method used is Support Vector Machine (SVM). The main goal of this research work is to help the medical field with a light of image processing.
We aim to separate the generative factors of data into two latent vectors in a variational autoencoder. One vector captures class factors relevant to target classification tasks, while the other vector captures style factors relevant to the remaining information. To learn the discrete class features, we introduce supervision using a small amount of labeled data, which can simply yet effectively reduce the effort required for hyperparameter tuning performed in existing unsupervised methods. Furthermore, we introduce a learning objective to encourage statistical independence between the vectors. We show that (i) this vector independence term exists within the result obtained on decomposing the evidence lower bound with multiple latent vectors, and (ii) encouraging such independence along with reducing the total correlation within the vectors enhances disentanglement performance. Experiments conducted on several image datasets demonstrate that the disentanglement achieved via our method can improve classification performance and generation controllability.
Conventionally, AI models are thought to trade off explainability for lower accuracy. We develop a training strategy that not only leads to a more explainable AI system for object classification, but as a consequence, suffers no perceptible accuracy degradation. Explanations are defined as regions of visual evidence upon which a deep classification network makes a decision. This is represented in the form of a saliency map conveying how much each pixel contributed to the network's decision. Our training strategy enforces a periodic saliency-based feedback to encourage the model to focus on the image regions that directly correspond to the ground-truth object. We quantify explainability using an automated metric, and using human judgement. We propose explainability as a means for bridging the visual-semantic gap between different domains where model explanations are used as a means of disentagling domain specific information from otherwise relevant features. We demonstrate that this leads to improved generalization to new domains without hindering performance on the original domain.
We consider the problem of classifying a medical image dataset when we have a limited amounts of labels. This is very common yet challenging setting as labelled data is expensive, time consuming to collect and may require expert knowledge. The current classification go-to of deep supervised learning is unable to cope with such a problem setup. However, using semi-supervised learning, one can produce accurate classifications using a significantly reduced amount of labelled data. Therefore, semi-supervised learning is perfectly suited for medical image classification. However, there has almost been no uptake of semi-supervised methods in the medical domain. In this work, we propose an all-in-one framework for deep semi-supervised classification focusing on graph based approaches, which up to our knowledge it is the first time that an approach with minimal labels has been shown to such an unprecedented scale with medical data. We introduce the concept of hybrid models by defining a classifier as a combination between an energy-based model and a deep net. Our energy functional is built on the Dirichlet energy based on the graph p-Laplacian. Our framework includes energies based on the $\ell_1$ and $\ell_2$ norms. We then connected this energy model to a deep net to generate a much richer feature space to construct a stronger graph. Our framework can be set to be adapted to any complex dataset. We demonstrate, through extensive numerical comparisons, that our approach readily compete with fully-supervised state-of-the-art techniques for the applications of Malaria Cells, Mammograms and Chest X-ray classification whilst using only 20% of labels.
In this paper we challenge the common assumption that convolutional layers in modern CNNs are translation invariant. We show that CNNs can and will exploit the absolute spatial location by learning filters that respond exclusively to particular absolute locations by exploiting image boundary effects. Because modern CNNs filters have a huge receptive field, these boundary effects operate even far from the image boundary, allowing the network to exploit absolute spatial location all over the image. We give a simple solution to remove spatial location encoding which improves translation invariance and thus gives a stronger visual inductive bias which particularly benefits small data sets. We broadly demonstrate these benefits on several architectures and various applications such as image classification, patch matching, and two video classification datasets.
Complex assignments typically consist of open-ended questions with large and diverse content in the context of both classroom and online graduate programs. With the sheer scale of these programs comes a variety of problems in peer and expert feedback, including rogue reviews. As such with the hope of identifying important contents needed for the review, in this work we present a very first work on key phrase classification with a detailed empirical study on traditional and most recent language modeling approaches. From this study, we find that the task of classification of key phrases is ambiguous at a human level producing Cohen's kappa of 0.77 on a new data set. Both pretrained language models and simple TFIDF SVM classifiers produce similar results with a former producing average of 0.6 F1 higher than the latter. We finally derive practical advice from our extensive empirical and model interpretability results for those interested in key phrase classification from educational reports in the future.
We propose a simple definition of an explanation for the outcome of a classifier based on concepts from causality. We compare it with previously proposed notions of explanation, and study their complexity. We conduct an experimental evaluation with two real datasets from the financial domain.
Aspect-based opinion mining is the task of identifying sentiment at the aspect level in opinionated text, which consists of two subtasks: aspect category extraction and sentiment polarity classification. While aspect category extraction aims to detect and categorize opinion targets such as product features, sentiment polarity classification assigns a sentiment label, i.e. positive, negative, or neutral, to each identified aspect. Supervised learning methods have been shown to deliver better accuracy for this task but they require labeled data, which is costly to obtain, especially for resource-poor languages like Vietnamese. To address this problem, we present a supervised aspect-based opinion mining method that utilizes labeled data from a foreign language (English in this case), which is translated to Vietnamese by an automated translation tool (Google Translate). Because aspects and opinions in different languages may be expressed by different words, we propose using word embeddings, in addition to other features, to reduce the vocabulary difference between the original and translated texts, thus improving the effectiveness of aspect category extraction and sentiment polarity classification processes. We also introduce an annotated corpus of aspect categories and sentiment polarities extracted from restaurant reviews in Vietnamese, and conduct a series of experiments on the corpus. Experimental results demonstrate the effectiveness of the proposed approach.
Lung Cancer is the most common cause of cancer-related death worldwide. Early and automatic diagnosis of Solitary Pulmonary Nodules (SPN) in Computer Tomography (CT) chest scans can provide early treatment as well as doctor liberation from time-consuming procedures. Deep Learning has been proven as a popular and influential method in many medical imaging diagnosis areas. In this study, we consider the problem of diagnostic classification between benign and malignant lung nodules in CT images derived from a PET/CT scanner. More specifically, we intend to develop experimental Convolutional Neural Network (CNN) architectures and conduct experiments, by tuning their parameters, to investigate their behavior, and to define the optimal setup for the accurate classification. For the experiments, we utilize PET/CT images obtained from the Laboratory of Nuclear Medicine of the University of Patras, and the publically available database called Lung Image Database Consortium Image Collection (LIDC-IDRI). Furthermore, we apply simple data augmentation to generate new instances and to inspect the performance of the developed networks. Classification accuracy of 91% and 93% on the PET/CT dataset and on a selection of nodule images form the LIDC-IDRI dataset, is achieved accordingly. The results demonstrate that CNNs are a trustworth method for nodule classification. Also, the experiment confirms that data augmentation enhances the robustness of the CNNs.
Few-shot detection and classification have advanced significantly in recent years. Yet, detection approaches require strong annotation (bounding boxes) both for pre-training and for adaptation to novel classes, and classification approaches rarely provide localization of objects in the scene. In this paper, we introduce StarNet - a few-shot model featuring an end-to-end differentiable non-parametric star-model detection and classification head. Through this head, the backbone is meta-trained using only image-level labels to produce good features for jointly localizing and classifying previously unseen categories of few-shot test tasks using a star-model that geometrically matches between the query and support images (to find corresponding object instances). Being a few-shot detector, StarNet does not require any bounding box annotations, neither during pre-training nor for novel classes adaptation. It can thus be applied to the previously unexplored and challenging task of Weakly Supervised Few-Shot Object Detection (WS-FSOD), where it attains significant improvements over the baselines. In addition, StarNet shows significant gains on few-shot classification benchmarks that are less cropped around the objects (where object localization is key).
Datasets with noisy labels are a common occurrence in practical applications of classification methods. We propose a simple probabilistic method for training deep classifiers under input-dependent (heteroscedastic) label noise. We assume an underlying heteroscedastic generative process for noisy labels. To make gradient based training feasible we use a temperature parameterized softmax as a smooth approximation to the assumed generative process. We illustrate that the softmax temperature controls a bias-variance trade-off for the approximation. By tuning the softmax temperature, we improve accuracy, log-likelihood and calibration on both image classification benchmarks with controlled label noise as well as Imagenet-21k which has naturally occurring label noise. For image segmentation, our method increases the mean IoU on the PASCAL VOC and Cityscapes datasets by more than 1% over the state-of-the-art model.