United States

We introduce equi-tuning, a novel fine-tuning method that transforms (potentially non-equivariant) pretrained models into group equivariant models while incurring minimum $L_2$ loss between the feature representations of the pretrained and the equivariant models. Large pretrained models can be equi-tuned for different groups to satisfy the needs of various downstream tasks. Equi-tuned models benefit from both group equivariance as an inductive bias and semantic priors from pretrained models. We provide applications of equi-tuning on three different tasks: image classification, compositional generalization in language, and fairness in natural language generation (NLG). We also provide a novel group-theoretic definition for fairness in NLG. The effectiveness of this definition is shown by testing it against a standard empirical method of fairness in NLG. We provide experimental results for equi-tuning using a variety of pretrained models: Alexnet, Resnet, VGG, and Densenet for image classification; RNNs, GRUs, and LSTMs for compositional generalization; and GPT2 for fairness in NLG. We test these models on benchmark datasets across all considered tasks to show the generality and effectiveness of the proposed method.

AAAI 2023

Equi-Tuning: Group Equivariant Fine-Tuning of Pretrained Models

group equivariance

transfer-learning

image classification

natural language generation

fairness

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-23 is the Thirty-Seventh AAAI Conference on Artificial Intelligence. The theme of this conference is to create collaborative bridges within and beyond AI. Like previous AAAI conferences, AAAI-23 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and two new activities: a Bridge Program and a Lab Program. Many of these activities are tailored to the theme of bridges and all are selected according to the highest standards, with additional programs for students and young researchers. 
AAAI is providing you with a conference planner, which you can use to help organize your itinerary of activities. This includes talks to attend in person, talks to attend remotely, breaks with colleagues and your site seeing activities. To access this conference planner, please go to [https://aaai-2023.takemobi.io](https://aaai-2023.takemobi.io).

In order to access this site, you need to register. If you haven't already, please register [here](https://aaai.org/Conferences/AAAI-23/registration/).


The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines.

poster

We propose ordering-based approaches for learning the maximal ancestral graph (MAG) of a structural equation model (SEM) up to its Markov equivalence class (MEC) in the presence of unobserved variables. Existing ordering-based methods in the literature recover a graph through learning a causal order (c-order). We advocate for a novel order called removable order (r-order) as they are advantageous over c-orders for structure learning. This is because r-orders are the minimizers of an appropriately defined optimization problem that could be either solved exactly (using a reinforcement learning approach) or approximately (using a hill-climbing search). Moreover, the r-orders (unlike c-orders) are invariant among all the graphs in a MEC and include c-orders as a subset. Given that set of r-orders is often significantly larger than the set of c-orders, it is easier for the optimization problem to find an r-order instead of a c-order. We evaluate the performance and the scalability of our proposed approaches on both real-world and randomly generated networks.

Novel Ordering-based Approaches for Causal Structure Learning in the Presence of Unobserved Variables

  Herein, we propose a Spearman rank correlation-based screening procedure for ultrahigh-dimensional data with censored response cases. The proposed method is model-free without specifying any regression forms of predictors or response variables and is robust under the unknown monotone transformations of these response variable and predictors. The sure-screening and rank-consistency properties are established under some mild regularity conditions. Simulation studies demonstrate that the new screening method performs well in the presence
of a heavy-tailed distribution, strongly dependent predictors or outliers, and offers superior performance over the existing nonparametric screening procedures. In particular, the new screening method still works well when a response variable is observed under a high censoring rate. An illustrative example is provided.

Spearman Rank Correlation Screening for Ultrahigh-dimensional Censored Data

Determining kernels' sizes of a CNN model is a crucial and non-trivial design choice and has a significant impact on its performance. The majority of kernel size design methods rely on complex heuristic tricks or leverage neural architecture search that requires extreme computational resources. Thus, learning kernels' sizes jointly with the classifier's weights has been proposed as a workaround. However, the proposed methods cannot achieve satisfactory results or are inefficient for large-scale datasets. To fill in this gap, we design a novel efficient kernel size learning method in which a size predictor model learns to predict optimal kernels' sizes for a classifier given a desired number of parameters. It does so in collaboration with a hypernetwork that predicts the weights of the classifier's kernels - given kernel sizes predicted by the kernel predictor - to minimize the training objective and both models are trained end-to-end. Our method only needs a small fraction of training epochs of the original classifier to train these two models and find proper kernel sizes for it. Thus, it offers an efficient and effective solution for the kernel size learning problem. Our extensive experiments on MNIST, CIFAR-10, STL-10, and ImageNet-32 clearly demonstrates that our method can achieve the best training time vs. accuracy trade-off compared to previous kernel size learning methods and significantly outperform them on challenging datasets such as STL-10 and ImageNet-32.

EffConv: Efficient Learning of Kernel Sizes for Convolution Layers of CNNs

Due to the more complicated population dynamics of the NSGA-II, none of the existing runtime guarantees for this algorithm is accompanied by a non-trivial lower bound. Via a first mathematical understanding of the population dynamics of the NSGA-II, that is, by estimating the expected number of individuals having a certain objective value, we prove that the NSGA-II with suitable population size needs $\Omega(Nn\log n)$ function evaluations to find the Pareto front of the \oneminmax problem and $\Omega(Nn^k)$  evaluations on the \ojzj problem with jump size~$k$. These bounds are asymptotically tight (that is, they match previously shown upper bounds) and show that the NSGA-II here does not even in terms of the parallel runtime (number of iterations) profit from larger population sizes. For the \ojzj problem and when the same sorting is used for the computation of the crowding distance contributions of the two objectives, we even obtain a runtime estimate that is tight including the leading constant.

From Understanding the Population Dynamics of the NSGA-II to the First Proven Lower Bounds

Fairness has been considered as a critical problem in federated learning (FL). In this work, we analyze two direct causes of unfairness in FL - an unfair direction and an improper step size when updating the model. To solve these issues, we introduce an effective way to measure fairness of the model through the cosine similarity, and then propose a federated multiple gradient descent algorithm with fair guidance (FedMDFG) to drive the model fairer. We first convert FL into a multi-objective optimization problem (MOP) and design an advanced multiple gradient descent algorithm to calculate a fair descent direction by adding a fair-driven objective to MOP. A low-communication-cost line search strategy is then designed to find a better step size for the model update. We further show the theoretical analysis on how it can enhance fairness and guarantee the convergence. Finally, extensive experiments in several FL scenarios verify that FedMDFG is robust and outperforms the SOTA FL algorithms in convergence and fairness.

FedMDFG: Federated Learning with Multi-Gradient Descent and Fair Guidance

Recently, word enhancement has become very popular for Chinese Named Entity Recognition (NER), reducing segmentation errors and increasing the semantic and boundary information of Chinese words. However, these methods tend to ignore the semantic relationship before and after the sentence after integrating lexical information. Therefore, the regularity of word length information has not been fully explored in various word-character fusion methods. In this work, we propose a Lexicon-Attention and Data-Augmentation (LADA) method for Chinese NER. We discuss the challenges of using existing methods in incorporating word information for NER and show how our proposed methods could be leveraged to overcome those challenges. LADA is based on a Transformer Encoder that utilizes lexicon to construct a directed graph and fuses word information through updating the optimal edge of the graph. Specially, we introduce the advanced data augmentation method to obtain the optimal representation for the NER task. Experimental results show that the augmentation done using LADA can considerably boost the performance of our NER system and achieve significantly better results than previous state-of-the-art methods and variant models in the literature on four publicly available NER datasets, namely Resume, MSRA, Weibo, and OntoNotes v4. We also observe better generalization and application to a real-world setting from LADA on multi-source complex entities.

LADA-Trans-NER: Adaptive Efficient Transformer for Chinese Named Entity Recognition using Lexicon-Attention and Data-Augmentation

This paper revisits building  machine learning algorithms that involve interactions between entities, such as those between financial assets in an actively managed portfolio, or interactions between users in a social network. Our goal is to forecast the future evolution of ensembles of multivariate time series in such applications (e.g., the future return of a financial asset or the future popularity of a Twitter account). Designing ML algorithms for such systems requires addressing the challenges of high-dimensional interactions and non-linearity. Existing approaches usually adopt an ad-hoc approach to integrating high-dimensional techniques into non-linear models and recent studies have shown these approaches have questionable efficacy in time-evolving interacting systems. 


To this end, we propose a novel framework, which we dub as the additive influence model. Under our modeling assumption, we show that it is possible to decouple the learning of high-dimensional interactions from the learning of non-linear feature interactions. To learn the high-dimensional interactions, we leverage kernel-based techniques, with provable guarantees, to embed the entities in a low-dimensional latent space. To learn the non-linear feature-response interactions, we generalize prominent machine learning techniques, including designing a new statistically sound non-parametric method and an ensemble learning algorithm optimized for vector regressions. 
Extensive experiments on two common applications demonstrate that our new algorithms deliver significantly stronger forecasting power compared to standard and recently proposed methods. 

Symphony in the Latent Space: Provably Integrating High-dimensional Techniques with Non-linear Machine Learning Models

Human trajectory Prediction (HTP) in complex social environments plays a crucial and fundamental role in artificial intelligence systems. Conventional methods make use of both history behaviors and social interactions to forecast future trajectories. However, we demonstrate that the social environment is a confounder that misleads the model to learn spurious correlations between history and future trajectories. To end this, we first formulate the social environment, history and future trajectory variables into a structural causal model to analyze the causalities among them. Based on causal intervention rather than conventional likelihood, we propose a Social Environment ADjustment (SEAD) method, to remove the confounding effect of the social environment. The core of our method is implemented by a Social Cross Attention (SCA) module, which is universal, simple and effective. Our method has consistent improvements on ETH-UCY datasets with three baseline models and achieves competitive performances with existing methods.

Causal Intervention for Human Trajectory Prediction with Cross Attention Mechanism

Safe Interval Path Planning (SIPP) is a powerful algorithm for solving single-agent pathfinding problems when the agent is confined to a graph and certain vertices/edges of this graph are blocked at certain time intervals due to dynamic obstacles that populate the environment. Original SIPP algorithm relies on the assumption that the agent is able to stop instantaneously. However, this assumption often does not hold in practice, e.g. a mobile robot moving with a cruising speed is not able to stop immediately but rather requires gradual deceleration to a full stop that takes time. In other words, the robot is subject to kinodynamic constraints. Unfortunately, as we show in this work, in such a case original SIPP is incomplete. To this end, we introduce a novel variant of SIPP that is provably complete and optimal for planning with acceleration/deceleration. In the experimental evaluation we show that the key property of the original SIPP still holds for the modified version -- it performs much less expansions compared to A* and, as a result, is notably faster. 

Safe Interval Path Planning With Kinodynamic Constraints

Regression trees are one of the oldest forms of AI models, and their predictions can be made without a calculator, which makes them broadly useful, particularly for high-stakes applications. Within the large literature on regression trees, there has been little effort towards full provable optimization, mainly due to the computational hardness of the problem. This work proposes a dynamic programming-with-bounds approach to the construction of provably-optimal sparse regression trees. We leverage a novel lower bound based on an optimal solution to the k-Means clustering algorithm on one dimensional data. We are often able to find optimal sparse trees in seconds, even for challenging datasets that involve large numbers of samples and highly-correlated features.

Premium content

Equi-Tuning: Group Equivariant Fine-Tuning of Pretrained Models

Downloads

Next from AAAI 2023

Novel Ordering-based Approaches for Causal Structure Learning in the Presence of Unobserved Variables

Similar lecture

Learning from Training Dynamics: Identifying Mislabeled Data Beyond Manually Designed Features