Machine learning refactors knowledge to improve learning efficiency.
problem Inductive program synthesis efficiency through knowledge restructuring.
method Introduces Knorf, a system that refactors knowledge bases using constraint optimization.
result Learning from refactored knowledge improves predictive accuracy fourfold and reduces learning time by half.
Various problems in data analysis and statistical genetics call for recovery of a column-sparse, low-rank matrix from noisy observations. We propose ReFACTor, a simple variation of the classical Truncated Singular Value Decomposition (TSVD) algorithm. In contrast to previous sparse principal component analysis (PCA) al…
The reparameterization trick enables optimizing large scale stochastic computation graphs via gradient descent. The essence of the trick is to refactor each stochastic node into a differentiable function of its parameters and a random variable with fixed distribution. After refactoring, the gradients of the loss propag…
A modular framework for knowledge distillation simplifies experiments and reproducibility.
problem Difficulty in reproducing knowledge distillation studies due to lack of generalized frameworks.
method A configuration-driven PyTorch framework for knowledge distillation studies.
result Demonstrates efficient training strategies and various knowledge distillation methods.
funcGNN uses graph neural networks to estimate program similarity efficiently.
problem Estimating accurate program similarity for software engineering tasks.
method funcGNN trains on labeled CFG pairs to predict GED between unseen programs using effective embedding vectors.
result funcGNN achieves lower error rate (0.00194) and is 23 times faster than traditional methods.
Study uses supercomputers to improve financial predictions.
problem Improving financial predictions through better exploration of data.
method Refactored and ran algorithm on Fugaku supercomputer, exploring more rules.
result Increasing the number of explored rules improves predictive performance.
Neural program embedding can be helpful in analyzing large software, a task that is challenging for traditional logic-based program analyses due to their limited scalability. A key focus of recent machine-learning advances in this area is on modeling program semantics instead of just syntax. Unfortunately evaluating su…
COCA refactors training data to identify and erase unsafe concepts in LLMs.
problem Identifying and erasing unsafe concepts in Large Language Models (LLMs) for safety alignment.
method Concept Concentration (COCA) refactors training data with an explicit reasoning process to identify and erase unsafe concepts.
result COCA significantly reduces both in-distribution and out-of-distribution jailbreak success rates while maintaining strong performance on regular tasks.
Identifier names convey useful information about the intended semantics of code. Name-based program analyses use this information, e.g., to detect bugs, to predict types, and to improve the readability of code. At the core of name-based analyses are semantic representations of identifiers, e.g., in the form of learned …
fastai simplifies deep learning with a layered API.
problem Creating state-of-the-art deep learning models efficiently.
method Carefully layered architecture, Python type dispatch, GPU optimization, and novel callbacks.
result fastai achieves high performance with minimal compromise in usability and flexibility.
Recent student knowledge modeling algorithms such as Deep Knowledge Tracing (DKT) and Dynamic Key-Value Memory Networks (DKVMN) have been shown to produce accurate predictions of problem correctness within the same learning system. However, these algorithms do not attempt to directly infer student knowledge. In this pa…
A new framework for knowledge graph embedding using sheaves.
problem Learning representations for entities and relations in knowledge graphs.
method Using cellular sheaves to describe knowledge graph embeddings with consistency constraints.
result A generalized framework for reasoning about knowledge graph embedding models.
Framework evaluates the impact of prior knowledge in deep learning models.
problem Mitigating data-driven model shortcomings like data dependence and generalization ability.
method Model-agnostic framework inspired by interpretable machine learning, assessing data volume and estimation range effects.
result Complex relationship between data and knowledge, including dependence, synergistic, and substitution effects.
It is always demanding to learn robust visual representation for various learning problems; however, this learning and maintenance process usually suffers from noise, incompleteness or knowledge domain mismatch. Thus, robust representation learning by removing noisy features or samples, complementing incomplete data, a…
This paper aims to analyze knowledge consistency between pre-trained deep neural networks. We propose a generic definition for knowledge consistency between neural networks at different fuzziness levels. A task-agnostic method is designed to disentangle feature components, which represent the consistent knowledge, from…
Knowledge graph construction consists of two tasks: extracting information from external resources (knowledge population) and inferring missing information through a statistical analysis on the extracted information (knowledge completion). In many cases, insufficient external resources in the knowledge population hinde…
This paper enhances language models with knowledge awareness.
problem Understanding how much knowledge pretrained language models grasp.
method Inserting explicit knowledge layers into pretraining without changing transformer architecture.
result Significantly more knowledge packed into transformer parameters.
Survey of methods to incorporate external knowledge into stock price prediction.
problem Challenges in predicting stock prices due to market volatility and non-linearity.
method Survey of methods for acquiring and incorporating external knowledge into stock price prediction models.
result Systematic synthesis of previous studies on external knowledge types and their application in stock price prediction.
New model enables AI to learn autonomously.
problem Enabling AI to acquire domain knowledge.
method Hybrid model combining ontology, knowledge graph, and Logic Neural Network.
result System can enrich and extend its knowledge.
We formulate one methodology to put a value or price on knowledge using well accepted techniques from finance. We provide justifications for these finance principles based on the limitations of the physical world we live in. We start with the intuition for our method to value knowledge and then formalize this idea with…
This paper explores the connection between adversarial and knowledge transferability.
problem Understanding the factors affecting knowledge transferability.
method Theoretical analysis and practical metrics for adversarial transferability.
result Adversarial transferability and knowledge transferability are closely related.
Within Reinforcement Learning, there is a growing collection of research which aims to express all of an agent's knowledge of the world through predictions about sensation, behaviour, and time. This work can be seen not only as a collection of architectural proposals, but also as the beginnings of a theory of machine k…
Exploiting the appropriate inductive bias based on the knowledge of data is essential for achieving good performance in statistical machine learning. In practice, however, the domain knowledge of interest often provides information on the relationship of data attributes only distantly, which hinders direct utilization …
We present a baseline approach for cross-modal knowledge fusion. Different basic fusion methods are evaluated on existing embedding approaches to show the potential of joining knowledge about certain concepts across modalities in a fused concept representation.
Recent variants improve knowledge distillation performance.
problem Improving the performance of knowledge distillation.
method Introducing additional components or changing the learning process.
result These variants have shown promising results.
We present a family of novel methods for embedding knowledge graphs into real-valued tensors. These tensor-based embeddings capture the ordered relations that are typical in the knowledge graphs represented by semantic web languages like RDF. Unlike many previous models, our methods can easily use prior background know…
This paper improves understanding of knowledge distillation and its effects.
problem Improving model quality with a fixed capacity budget.
method Categorizing teacher's knowledge into three levels and studying their effects on distillation.
result The three hierarchical levels of teacher's knowledge (universe, domain, instance) significantly impact knowledge distillation.
The paper explains knowledge distillation by analyzing visual concepts in DNNs.
problem Understanding how knowledge distillation affects the learning of visual concepts in deep neural networks.
method The paper proposes three hypotheses and designs mathematical metrics to evaluate feature representations of DNNs.
result The hypotheses were verified through experiments on various DNNs.
In this paper, we consider a supervised learning setting where side knowledge is provided about the labels of unlabeled examples. The side knowledge has the effect of reducing the hypothesis space, leading to tighter generalization bounds, and thus possibly better generalization. We consider several types of side knowl…
Paper extends knowledge gradient for preferential BO, overcoming computational challenges.
problem Extending knowledge gradient to preferential BO with pairwise comparisons.
method Derive exact and analytical knowledge gradient for preferential BO.
result Exact knowledge gradient outperforms existing acquisition functions on benchmark problems.
New model of vague knowledge without strict partitions or transitivity.
problem Standard economic models of information fail to capture real-world vague knowledge.
method Relaxing assumptions of transitivity and partition structure to formalize vague knowledge.
result Vague knowledge can distinguish some states but not partition the state space.
b-LOAD extends local causal discovery with prior knowledge, improving causal effect estimation.
problem Local causal discovery struggles in data-scarce settings due to uncertainty and incomplete neighborhoods.
method b-LOAD incorporates prior knowledge directly into local structure learning, using Meek's rules to refine discovery.
result b-LOAD refines the admissible equivalence class and enlarges identifiable causal queries, improving causal effect estimation.
Bayesian neural networks incorporate domain knowledge through variational inference.
problem Specifying priors for Bayesian neural networks that capture domain knowledge is challenging.
method Proposes a framework for integrating domain knowledge into BNN priors through variational inference.
result BNNs with proposed domain knowledge priors outperform those with standard priors, achieving better predictive performance.
This work integrates domain knowledge into A*-based causal discovery methods.
problem Efficiently incorporating domain knowledge into A*-based causal discovery methods.
method Integrates various types of domain knowledge into A*-based causal discovery methods, reducing the graph search space and improving computational gains.
result Small amounts of domain knowledge can dramatically speed up A*-based causal discovery and improve its performance and practicality.
Proposes a new method for predicting missing relations in knowledge graphs.
problem Predicting missing relations between entities in knowledge graphs.
method Relational message passing method considering only edge features without entity IDs.
result PathCon method outperforms state-of-the-art methods significantly.
RID framework quantifies and regularizes task-relevant knowledge in distillation.
problem Distilling irrelevant information can hinder student model performance.
method Partial Information Decomposition to quantify and regularize task-relevant knowledge.
result RID framework leads to more resilient distillation under nuisance teachers.
New framework uses background knowledge to speed up causal discovery.
problem Scalable causal discovery for large datasets.
method Utilizes background knowledge during causal discovery process.
result Background knowledge reduces computational requirements and improves structure quality.
Role-wise data augmentation improves knowledge distillation effectiveness.
problem Existing knowledge distillation methods fail to utilize the full potential of teacher-student data interaction.
method Design and implement data augmentation agents with distinct roles for teacher and student.
result Specially tailored data points enhance the demonstration of teacher's knowledge to the student.
Knowledge distillation (KD) is a technique to derive optimal performance from a small student network (SN) by distilling knowledge of a large teacher network (TN) and transferring the distilled knowledge to the small SN. Since a role of convolutional neural network (CNN) in KD is to embed a dataset so as to perform a g…
Semantic TrueLearn uses semantic graphs to improve educational recommendation systems.
problem Challenges in handling semantic and hierarchical structure in knowledge areas.
method Introduces a novel learner model that exploits semantic relatedness between knowledge components using a Wikipedia link graph.
result Achieves statistically significant improvements in predictive performance for educational engagement.
RKD improves model compression by distilling residual knowledge from a deep teacher model.
problem Performance degradation due to the gap between student and teacher models.
method Introduces an assistant model to distill residual knowledge from the teacher model.
result RKD achieves better results on popular classification datasets than state-of-the-art methods.
Cluster-DAGs improve causal discovery with prior knowledge.
problem Finding cause-effect relationships from high-dimensional data.
method Cluster-DAGs as prior knowledge framework, modified constraint-based algorithms Cluster-PC and Cluster-FCI.
result Cluster-PC and Cluster-FCI outperform baselines without prior knowledge.
KGRL uses reinforcement learning with knowledge graphs for better interactive recommendation.
problem Achieving responsiveness and accuracy in dynamic user-item interactions.
method KGRL combines reinforcement learning and knowledge graphs, using a local knowledge network and attention mechanism.
result KGRL outperforms state-of-the-art methods in simulated and real-world environments.
Link prediction in large knowledge graphs has received a lot of attention recently because of its importance for inferring missing relations and for completing and improving noisily extracted knowledge graphs. Over the years a number of machine learning researchers have presented various models for predicting the prese…
Domain knowledge helps detect adversarial examples in multi-label classification.
problem Detecting adversarial examples in multi-label classification.
method Convert domain knowledge into constraints and inject them into a semi-supervised learning problem.
result Domain-knowledge constraints help detect adversarial examples effectively.
Paper proposes KE-GCN for better graph embedding.
problem Efficiently leveraging complex graphs with heterogeneous entities and relations.
method Combines GCNs and knowledge embedding methods.
result Advantages in knowledge graph alignment and entity classification.
Study on teaching with imperfect knowledge, showing its impact on optimal teaching sets.
problem Effect of imperfect teacher knowledge on effective teaching.
method Connections to machine teaching problem, optimal teaching sets.
result Teacher's success or failure depends on imperfect knowledge.
We report a neural architecture search framework, BioNAS, that is tailored for biomedical researchers to easily build, evaluate, and uncover novel knowledge from interpretable deep learning models. The introduction of knowledge dissimilarity functions in BioNAS enables the joint optimization of predictive power and bio…