An Exploration into why Output Regularization Mitigates Label Noise

Label noise presents a real challenge for supervised learning algorithms . Noise robust losses is one of the more promising approaches for dealing with label noise . But their ability to mitigate label noise lackmathematical rigor. In this work we aim at closing this gap by showing thatlosses, which incorporate an output regularization term, become symmetric .…

DABT A Dependency aware Bug Triaging Method

The Dependency-aware Bug Triaging (DABT) leverages natural language processing and integer programming to assign bugs to developers . DABT is able to reduce the number of overdue bugs up to 12\% . It also decreases the average fixing time of the bugs by half .…

A deep learning model for gastric diffuse type adenocarcinoma classification in whole slide images

Gastric diffuse-type adenocarcinoma represents a disproportionately highpercentage of cases of gastric cancers occurring in the young . Usually it affects the body of the stomach,and presents shorter duration and worse prognosis compared with thedifferentiated (intestinal) type of adenodecarinoma . The main difficultyencountered in the differential diagnosis occurs with the diffuse type.…

PanGu α Large scale Autoregressive Pretrained Chinese Language Models with Auto parallel Computation

Large-scale Pretrained Language Models (PLMs) have become the new paradigm for Natural Language Processing (NLP) PLMs with hundreds of billionsparameters such as GPT-3 have demonstrated strong performances on naturallanguage understanding and generation with \textit{few-shot in-context}learning . In this work, we present our practice on training large-scaleautoregressive language models named PanGu-$\alpha$ with up to 200 billionparameters .…

TrustyAI Explainability Toolkit

Artificial intelligence (AI) is becoming increasingly more popular and can be found in workplaces and homes around the world . Regulation changes such as the GDPR mean that users have aright to understand how their data has been processed and saved .…

Boolean Reasoning Based Biclustering for Shifting Pattern Extraction

Shifting patterns are interesting as they account constantfluctuations in data, i.e. they capture situations in which all the values inthe pattern move up or down for one dimension maintaining the range amplitudefor all the dimensions . The induction of shifting patterns by means of Boolean reasoning is due to theability of finding all inclusion–maximal {\delta-shifting patterns .…

Diverse Image Inpainting with Bidirectional and Autoregressive Transformers

Image inpainting is an underdetermined inverse problem, it naturally allowsdiverse contents that fill up missing or corrupted regions reasonably andrealistically . Prevalent approaches using convolutional neural networks (CNNs) suffer from limited perception fields for capturing global features . We propose BAT-Fill, an image inpaining framework with a novel bidirectionalautoregressive transformer (BAT-Fill) that models deep bidsirectional contexts for the generation of diverse contents .…

An Algorithm to Effect Prompt Termination of Myopic Local Search on Kauffman s NK Landscape

In the NK model given by Kauffman, myopic local search involves flipping onerandomly-chosen bit of an N-bit decision string in every time step and accepting the new configuration if that has higher fitness . This algorithm consumes the full extent of computational resources allocated – given by the number of alternative configurations inspected – even though search is expected to terminate the moment there are no neighbors having higher fitness.…

2 5D Visual Relationship Detection

Visual 2.5D perception involves understanding the semantics and geometry of ascene through reasoning about object relationships with respect to the viewer in an environment . Unlike general VRD, 2 .5VRD isegocentric, using the camera’s viewpoint as a common reference . Unlike depth estimation, it is object-centric and not onlyfocuses on depth .…

EigenGAN Layer Wise Eigen Learning for GANs

EigenGAN is able to unsupervisedly mine interpretable and controllable dimensions from different generator layers . The algorithm can discover controlled dimensions for high-level concepts such as pose and gender in the subspace of deep layers, as well as low-level terms such as hue and color .…

Visformer The Vision friendly Transformer

Visformer outperforms both theTransformer-based and convolution-based models in terms of ImageNetclassification accuracy . The advantage becomes more significant when the training set is smaller . The code is available at https://://://github.com/danczs/Visformer. It is abbreviated from the `Vision-friendly Transformer’ with the same computational complexity as the Visformer architecture .…

Towards Knowledge Graphs Validation through Weighted Knowledge Sources

The performance of applications rely on high-quality knowledge bases, a.k.a. Knowledge Graphs . To ensure their quality one important task is KnowledgeValidation, which measures the degree to which statements or triples of a Knowledge Graph (KG) are correct . We propose and implement a validation approach that computes a confidence score for everytriple and instance in a KG .…

Represent Items by Items An Enhanced Representation of the Target Item for Recommendation

Item-based collaborative filtering (ICF) has been widely used in industrial applications such as recommender system and online advertising . In this paper, we propose an enhanced representation of the target item which distills relevant information from the co-occurrence items . With the enhanced representation, CER has strongerrepresentation power for the tail items compared to the state-of-the-art ICFmethods.…

Bijective proofs for Eulerian numbers in types B and D

We give bijective proofs of the identity of Stembridge’s identity . We also establish abijective correspondence between even signed permutations and pairs of pairs of $w, E)$ with $([n, E), a threshold graph and $w$ a degree ordering of $($n), $(w) and $(W)$ (W)) The bijectives rely on a representation ofsigned permutations as paths .…

Algorithmic Solution for Non Square Dense Systems of Linear Equations with applications in Feature Selection

We present a novel algorithm attaining excessively fast, the sought solutionof linear systems of equations . The execution time is very short compared with state-of-the-art methods, exhibiting up to $10^3 speed-up and lowmemory allocation demands . The accuracy is high and straightforwardly controlled, and the numerical results highlight the efficiency of the proposed algorithm, in terms of computation time, solution accuracy and solution accuracy .…

Recalibration of Aleatoric and Epistemic Regression Uncertainty in Medical Imaging

The consideration of predictive uncertainty in medical imaging with deeplearning is of utmost importance . We apply $ \sigma $ scaling with a single scalarvalue; a simple, yet effective calibration method for both types ofuncertainty . The performance of our approach is evaluated on a variety of common medical regression data sets using different state-of-the-artconvolutional network architectures .…

Reconfigurable Adaptive Channel Sensing

Channel sensing consists of probing the channel from time to time to check whether or not it is active – say of an incoming message . When communication is sparse with information being sent once in a long while,channel sensing becomes a significant source of energy consumption .…

Non Parametric Few Shot Learning for Word Sense Disambiguation

Word sense disambiguation (WSD) is a long-standing problem in naturallanguage processing . A significant challenge in supervised all-words WSD is to classify among senses for a majority of words that lie in the long-taildistribution . In this work, we proposeMetricWSD, a non-parametric few-shot learning approach to mitigate this dataimbalance issue .…

A unified Neural Network Approach to E CommerceRelevance Learning

Result relevance scoring is critical to e-commerce search user experience . Traditional information retrieval methods focus on keyword matching and counting-based numeric features . We describe a feed-forward neuralmodel to provide relevance score for (query, item) pairs . We found significantimprovement over GBDT baseline as well as several off-the-shelf deep-learningbaselines on an independently constructed ratings dataset .…

Speeding up Computational Morphogenesis with Online Neural Synthetic Gradients

A wide range of modern science and engineering applications are formulated as optimization problems with partial differential equations (PDEs) These PDE-constrained optimization problems are typically solved using a standard discretize-then-optimize approach . In many industry applicationsthat require high-resolution solutions, the discretized constraints can easilyhave millions or even billions of variables, making it very slow for the standard iterative optimizer to solve the exact gradients .…

Causal Learning for Socially Responsible AI

There have been increasing concerns about Artificial Intelligence due to its unfathomable potential power . Researchers proposed to develop socially responsible AI (SRAI) One of these approaches is causal learning (CL) We survey state-of-the-art methods of CL for SRAI . We begin by examining the seven CL tools to enhance the social responsibility of AI .…

A Comprehensive Attempt to Research Statement Generation

Research statement generation (RSG) task aims to summarize one’s researchachievements and help prepare a formal research statement . For this task, we construct an RSG dataset with 62 research statements and the corresponding 1,203 publications . We propose a practical RSG method which identifies arearcher’s research directions by topic modeling and clustering techniques .…

Math Operation Embeddings for Open ended Solution Analysis and Feedback

Feedback on student answers and even during intermediate steps in solving questions is an important element in math education . Such feedback can help students correct their errors and ultimately lead toimproved learning outcomes . Most existing approaches for automated studentsolution analysis and feedback require manually constructing cognitive models and anticipating student errors for each question .…

An Adaptive Learning based Generative Adversarial Network for One To One Voice Conversion

Voice Conversion (VC) deals with conversion of vocal style of one speaker to another speaker while keeping the linguistic contentsunchanged . VC task is performed through a three-stage pipeline consisting of speech analysis, speech feature mapping, and speech reconstruction . ALGAN-VC framework consists of some approaches to improve speech quality and voice similarity between source and target speakers .…