-
Li-Yau type and Harnack estimates for systems of reaction-diffusion equations via hybrid curvature-dimension condition
Authors:
Sebastian Kräss,
Rico Zacher
Abstract:
We prove Li-Yau and Harnack inequalities for systems of linear reaction-diffusion equations. By introducing an additional discrete spatial variable, the system is rewritten as a scalar diffusion equation with an operator sum. For such operators in a mixed continuous and discrete setting, we introduce the hybrid curvature-dimension condition $CD_{hyb} (κ,d)$, which is a combination of the Bakry-Éme…
▽ More
We prove Li-Yau and Harnack inequalities for systems of linear reaction-diffusion equations. By introducing an additional discrete spatial variable, the system is rewritten as a scalar diffusion equation with an operator sum. For such operators in a mixed continuous and discrete setting, we introduce the hybrid curvature-dimension condition $CD_{hyb} (κ,d)$, which is a combination of the Bakry-Émery condition $CD(κ,d)$ and one of its discrete analogues, the condition $CD_Υ(κ,d)$. We establish a hybrid tensorisation principle and prove that under $CD_{hyb} (0,d)$ with $d<\infty$ a differential Harnack estimate of Li-Yau type holds, from which a Harnack inequality can be deduced by an integration argument.
△ Less
Submitted 19 December, 2023;
originally announced December 2023.
-
Aronson-Bénilan and Harnack estimates for the discrete porous medium equation
Authors:
Sebastian Kräss,
Rico Zacher
Abstract:
We consider the porous medium equation (PME) on a locally finite graph and identify suitable curvature-dimension (CD) conditions under which a discrete version of the fundamental Aronson-Bénilan estimate holds true for positive solutions of the PME. We also show that these estimates allow to prove Harnack inequalities which are structurally similar to the continuous case. The new CD conditions are…
▽ More
We consider the porous medium equation (PME) on a locally finite graph and identify suitable curvature-dimension (CD) conditions under which a discrete version of the fundamental Aronson-Bénilan estimate holds true for positive solutions of the PME. We also show that these estimates allow to prove Harnack inequalities which are structurally similar to the continuous case. The new CD conditions are illustrated with several concrete examples, e.g.\ complete and chain-like graphs.
△ Less
Submitted 18 January, 2023;
originally announced January 2023.
-
Pile of Law: Learning Responsible Data Filtering from the Law and a 256GB Open-Source Legal Dataset
Authors:
Peter Henderson,
Mark S. Krass,
Lucia Zheng,
Neel Guha,
Christopher D. Manning,
Dan Jurafsky,
Daniel E. Ho
Abstract:
One concern with the rise of large language models lies with their potential for significant harm, particularly from pretraining on biased, obscene, copyrighted, and private information. Emerging ethical approaches have attempted to filter pretraining material, but such approaches have been ad hoc and failed to take context into account. We offer an approach to filtering grounded in law, which has…
▽ More
One concern with the rise of large language models lies with their potential for significant harm, particularly from pretraining on biased, obscene, copyrighted, and private information. Emerging ethical approaches have attempted to filter pretraining material, but such approaches have been ad hoc and failed to take context into account. We offer an approach to filtering grounded in law, which has directly addressed the tradeoffs in filtering material. First, we gather and make available the Pile of Law, a 256GB (and growing) dataset of open-source English-language legal and administrative data, covering court opinions, contracts, administrative rules, and legislative records. Pretraining on the Pile of Law may help with legal tasks that have the promise to improve access to justice. Second, we distill the legal norms that governments have developed to constrain the inclusion of toxic or private content into actionable lessons for researchers and discuss how our dataset reflects these norms. Third, we show how the Pile of Law offers researchers the opportunity to learn such filtering rules directly from the data, providing an exciting new research direction in model-based processing.
△ Less
Submitted 29 November, 2022; v1 submitted 1 July, 2022;
originally announced July 2022.
-
Li-Yau and Harnack inequalities via curvature-dimension conditions for discrete long-range jump operators including the fractional discrete Laplacian
Authors:
Sebastian Kräss,
Frederic Weber,
Rico Zacher
Abstract:
We consider operators of the form $L u(x) = \sum_{y \in \mathbb{Z}} k(x-y) \big( u(y) - u(x)\big)$ on the one-dimensional lattice with symmetric, integrable kernel $k$. We prove several results stating that under certain conditions on the kernel the operator $L$ satisfies the curvature-dimension condition $CD_Υ(0,F)$ (recently introduced by two of the authors) with some $CD$-function $F$, where at…
▽ More
We consider operators of the form $L u(x) = \sum_{y \in \mathbb{Z}} k(x-y) \big( u(y) - u(x)\big)$ on the one-dimensional lattice with symmetric, integrable kernel $k$. We prove several results stating that under certain conditions on the kernel the operator $L$ satisfies the curvature-dimension condition $CD_Υ(0,F)$ (recently introduced by two of the authors) with some $CD$-function $F$, where attention is also paid to the asymptotic properties of $F$ (exponential growth at infinity and power-type behaviour near zero). We show that $CD_Υ(0,F)$ implies a Li-Yau inequality for positive solutions of the heat equation associated with the operator $L$. The Li-Yau estimate in turn leads to a Harnack inequality, from which we also derive heat kernel bounds. Our results apply to a wide class of operators including the fractional discrete Laplacian.
△ Less
Submitted 12 January, 2022;
originally announced January 2022.
-
Context-Aware Legal Citation Recommendation using Deep Learning
Authors:
Zihan Huang,
Charles Low,
Mengqiu Teng,
Hongyi Zhang,
Daniel E. Ho,
Mark S. Krass,
Matthias Grabmair
Abstract:
Lawyers and judges spend a large amount of time researching the proper legal authority to cite while drafting decisions. In this paper, we develop a citation recommendation tool that can help improve efficiency in the process of opinion drafting. We train four types of machine learning models, including a citation-list based method (collaborative filtering) and three context-based methods (text si…
▽ More
Lawyers and judges spend a large amount of time researching the proper legal authority to cite while drafting decisions. In this paper, we develop a citation recommendation tool that can help improve efficiency in the process of opinion drafting. We train four types of machine learning models, including a citation-list based method (collaborative filtering) and three context-based methods (text similarity, BiLSTM and RoBERTa classifiers). Our experiments show that leveraging local textual context improves recommendation, and that deep neural models achieve decent performance. We show that non-deep text-based methods benefit from access to structured case metadata, but deep models only benefit from such access when predicting from context of insufficient length. We also find that, even after extensive training, RoBERTa does not outperform a recurrent neural model, despite its benefits of pretraining. Our behavior analysis of the RoBERTa model further shows that predictive performance is stable across time and citation classes.
△ Less
Submitted 20 June, 2021;
originally announced June 2021.
-
Automatic segmentation of the pulmonary lobes with a 3D u-net and optimized loss function
Authors:
Bianca Lassen-Schmidt,
Alessa Hering,
Stefan Krass,
Hans Meine
Abstract:
Fully-automatic lung lobe segmentation is challenging due to anatomical variations, pathologies, and incomplete fissures. We trained a 3D u-net for pulmonary lobe segmentation on 49 mainly publically available datasets and introduced a weighted Dice loss function to emphasize the lobar boundaries. To validate the performance of the proposed method we compared the results to two other methods. The…
▽ More
Fully-automatic lung lobe segmentation is challenging due to anatomical variations, pathologies, and incomplete fissures. We trained a 3D u-net for pulmonary lobe segmentation on 49 mainly publically available datasets and introduced a weighted Dice loss function to emphasize the lobar boundaries. To validate the performance of the proposed method we compared the results to two other methods. The new loss function improved the mean distance to 1.46 mm (compared to 2.08 mm for simple loss function without weighting).
△ Less
Submitted 29 May, 2020;
originally announced June 2020.
