-
End-to-end Offline Reinforcement Learning for Glycemia Control
Authors:
Tristan Beolet,
Alice Adenis,
Erik Huneker,
Maxime Louis
Abstract:
The development of closed-loop systems for glycemia control in type I diabetes relies heavily on simulated patients. Improving the performances and adaptability of these close-loops raises the risk of over-fitting the simulator. This may have dire consequences, especially in unusual cases which were not faithfully-if at all-captured by the simulator. To address this, we propose to use offline RL a…
▽ More
The development of closed-loop systems for glycemia control in type I diabetes relies heavily on simulated patients. Improving the performances and adaptability of these close-loops raises the risk of over-fitting the simulator. This may have dire consequences, especially in unusual cases which were not faithfully-if at all-captured by the simulator. To address this, we propose to use offline RL agents, trained on real patient data, to perform the glycemia control. To further improve the performances, we propose an end-to-end personalization pipeline, which leverages offline-policy evaluation methods to remove altogether the need of a simulator, while still enabling an estimation of clinically relevant metrics for diabetes.
△ Less
Submitted 16 October, 2023;
originally announced October 2023.
-
Puppeteer: A Random Forest-based Manager for Hardware Prefetchers across the Memory Hierarchy
Authors:
Furkan Eris,
Marcia S. Louis,
Kubra Eris,
Jose L. Abellan,
Ajay Joshi
Abstract:
Over the years, processor throughput has steadily increased. However, the memory throughput has not increased at the same rate, which has led to the memory wall problem in turn increasing the gap between effective and theoretical peak processor performance. To cope with this, there has been an abundance of work in the area of data/instruction prefetcher designs. Broadly, prefetchers predict future…
▽ More
Over the years, processor throughput has steadily increased. However, the memory throughput has not increased at the same rate, which has led to the memory wall problem in turn increasing the gap between effective and theoretical peak processor performance. To cope with this, there has been an abundance of work in the area of data/instruction prefetcher designs. Broadly, prefetchers predict future data/instruction address accesses and proactively fetch data/instructions in the memory hierarchy with the goal of lowering data/instruction access latency. To this end, one or more prefetchers are deployed at each level of the memory hierarchy, but typically, each prefetcher gets designed in isolation without comprehensively accounting for other prefetchers in the system. As a result, individual prefetchers do not always complement each other, and that leads to lower average performance gains and/or many negative outliers. In this work, we propose Puppeteer, which is a hardware prefetcher manager that uses a suite of random forest regressors to determine at runtime which prefetcher should be ON at each level in the memory hierarchy, such that the prefetchers complement each other and we reduce the data/instruction access latency. Compared to a design with no prefetchers, using Puppeteer we improve IPC by 46.0% in 1 Core (1C), 25.8% in 4 Core (4C), and 11.9% in 8 Core (8C) processors on average across traces generated from SPEC2017, SPEC2006, and Cloud suites with ~10KB overhead. Moreover, we also reduce the number of negative outliers by over 89%, and the performance loss of the worst-case negative outlier from 25% to only 5% compared to the state-of-the-art.
△ Less
Submitted 28 January, 2022;
originally announced January 2022.
-
Leveraging 3D Information in Unsupervised Brain MRI Segmentation
Authors:
Benjamin Lambert,
Maxime Louis,
Senan Doyle,
Florence Forbes,
Michel Dojat,
Alan Tucholka
Abstract:
Automatic segmentation of brain abnormalities is challenging, as they vary considerably from one pathology to another. Current methods are supervised and require numerous annotated images for each pathology, a strenuous task. To tackle anatomical variability, Unsupervised Anomaly Detection (UAD) methods are proposed, detecting anomalies as outliers of a healthy model learned using a Variational Au…
▽ More
Automatic segmentation of brain abnormalities is challenging, as they vary considerably from one pathology to another. Current methods are supervised and require numerous annotated images for each pathology, a strenuous task. To tackle anatomical variability, Unsupervised Anomaly Detection (UAD) methods are proposed, detecting anomalies as outliers of a healthy model learned using a Variational Autoencoder (VAE). Previous work on UAD adopted a 2D approach, meaning that MRIs are processed as a collection of independent slices. Yet, it does not fully exploit the spatial information contained in MRI. Here, we propose to perform UAD in a 3D fashion and compare 2D and 3D VAEs. As a side contribution, we present a new loss function guarantying a robust training. Learning is performed using a multicentric dataset of healthy brain MRIs, and segmentation performances are estimated on White-Matter Hyperintensities and tumors lesions. Experiments demonstrate the interest of 3D methods which outperform their 2D counterparts.
△ Less
Submitted 26 January, 2021;
originally announced January 2021.
-
Assessing the distribution of discrete survival time in presence of recall error
Authors:
Sedigheh Mirzaei Salehabadi,
Edwina Yeung,
Germaine M. Buck Louis,
Rajeshwari Sundaram
Abstract:
Retrospectively ascertained survival time may be subject to recall error. An example of discrete survival time with such recall error is time-to-pregnancy (TTP), the number of months non-contracepting couples require to get pregnant which is a measure of human fecundity. The epidemiological literature has demonstrated that retrospective TTP is subject to recall error and statistical models focusin…
▽ More
Retrospectively ascertained survival time may be subject to recall error. An example of discrete survival time with such recall error is time-to-pregnancy (TTP), the number of months non-contracepting couples require to get pregnant which is a measure of human fecundity. The epidemiological literature has demonstrated that retrospective TTP is subject to recall error and statistical models focusing on TTP have not accounted for the recall error. We propose a multistage model that utilizes women's retrospectively-reported TTP and associated certainty to estimate the TTP distribution. Our proposed model utilizes a discrete survival function that accounts for random heterogeneity arising from between women TTP data as well as a multinomial regression model to account for her certainty as accuracy may decline over time, i.e., depends on time since pregnancy in estimating the TTP distribution. Other novel features of the model include attention to whether the pregnancy was (un)planned as well as providing an approach to predict survival function for women without a reported TTP. Our model allows for the consideration of covariates for each of the underlying factors of (un)planned pregnancy, measure of certainty and TTP distribution. The proposed model is applicable for any discrete survival time when certainty in reporting may be a consideration. We use Monte Carlo simulations to assess the finite sample performance for the proposed estimators. We illustrate our proposed method using data from Upstate KIDS Study.
△ Less
Submitted 16 October, 2018;
originally announced October 2018.
-
Cautionary note on "Semiparametric modeling of grouped current duration data with preferential reporting'"
Authors:
Alexander C. McLain,
Rajeshwari Sundaram,
Marie Thoma,
Germaine M. Buck Louis
Abstract:
This report is designed to clarify a few points about the article "Semiparametric modeling of grouped current duration data with preferential reporting" by McLain, Sundaram, Thoma and Louis in Statistics in Medicine (McLain et al., 2014, hereafter MSTL) regarding using the methods under right censoring. In simulation studies, it has been found that bias can occur when right censoring is present. C…
▽ More
This report is designed to clarify a few points about the article "Semiparametric modeling of grouped current duration data with preferential reporting" by McLain, Sundaram, Thoma and Louis in Statistics in Medicine (McLain et al., 2014, hereafter MSTL) regarding using the methods under right censoring. In simulation studies, it has been found that bias can occur when right censoring is present. Current duration data normally does not have censored values, but censoring can be induced at a value, say tau, after which the data values are thought to be unreliable. As noted in MSTL, some right censored data require an assumption on the parametric form of the data beyond τ. While this assumption was given in MSTL, the implications of the assumption were not sufficiently explored. Here we present simulations and evaluate the methods of MSTL under type I censoring, give some settings under which the method works well even in presence of censoring, state when the model is correctly specified and discuss the reasons of the bias.
△ Less
Submitted 2 January, 2018;
originally announced January 2018.
-
Parallel transport in shape analysis: a scalable numerical scheme
Authors:
Maxime Louis,
Alexandre Bône,
Benjamin Charlier,
Stanley Durrleman
Abstract:
The analysis of manifold-valued data requires efficient tools from Riemannian geometry to cope with the computational complexity at stake. This complexity arises from the always-increasing dimension of the data, and the absence of closed-form expressions to basic operations such as the Riemannian logarithm. In this paper, we adapt a generic numerical scheme recently introduced for computing parall…
▽ More
The analysis of manifold-valued data requires efficient tools from Riemannian geometry to cope with the computational complexity at stake. This complexity arises from the always-increasing dimension of the data, and the absence of closed-form expressions to basic operations such as the Riemannian logarithm. In this paper, we adapt a generic numerical scheme recently introduced for computing parallel transport along geodesics in a Riemannian manifold to finite-dimensional manifolds of diffeomorphisms. We provide a qualitative and quantitative analysis of its behavior on high-dimensional manifolds, and investigate an application with the prediction of brain structures progression.
△ Less
Submitted 23 November, 2017;
originally announced November 2017.
-
Prediction of the progression of subcortical brain structures in Alzheimer's disease from baseline
Authors:
Alexandre Bône,
Maxime Louis,
Alexandre Routier,
Jorge Samper,
Michael Bacci,
Benjamin Charlier,
Olivier Colliot,
Stanley Durrleman
Abstract:
We propose a method to predict the subject-specific longitudinal progression of brain structures extracted from baseline MRI, and evaluate its performance on Alzheimer's disease data. The disease progression is modeled as a trajectory on a group of diffeomorphisms in the context of large deformation diffeomorphic metric mapping (LDDMM). We first exhibit the limited predictive abilities of geodesic…
▽ More
We propose a method to predict the subject-specific longitudinal progression of brain structures extracted from baseline MRI, and evaluate its performance on Alzheimer's disease data. The disease progression is modeled as a trajectory on a group of diffeomorphisms in the context of large deformation diffeomorphic metric mapping (LDDMM). We first exhibit the limited predictive abilities of geodesic regression extrapolation on this group. Building on the recent concept of parallel curves in shape manifolds, we then introduce a second predictive protocol which personalizes previously learned trajectories to new subjects, and investigate the relative performances of two parallel shifting paradigms. This design only requires the baseline imaging data. Finally, coefficients encoding the disease dynamics are obtained from longitudinal cognitive measurements for each subject, and exploited to refine our methodology which is demonstrated to successfully predict the follow-up visits.
△ Less
Submitted 23 November, 2017;
originally announced November 2017.
-
Field of Groves: An Energy-Efficient Random Forest
Authors:
Zafar Takhirov,
Joseph Wang,
Marcia S. Louis,
Venkatesh Saligrama,
Ajay Joshi
Abstract:
Machine Learning (ML) algorithms, like Convolutional Neural Networks (CNN), Support Vector Machines (SVM), etc. have become widespread and can achieve high statistical performance. However their accuracy decreases significantly in energy-constrained mobile and embedded systems space, where all computations need to be completed under a tight energy budget. In this work, we present a field of groves…
▽ More
Machine Learning (ML) algorithms, like Convolutional Neural Networks (CNN), Support Vector Machines (SVM), etc. have become widespread and can achieve high statistical performance. However their accuracy decreases significantly in energy-constrained mobile and embedded systems space, where all computations need to be completed under a tight energy budget. In this work, we present a field of groves (FoG) implementation of random forests (RF) that achieves an accuracy comparable to CNNs and SVMs under tight energy budgets. Evaluation of the FoG shows that at comparable accuracy it consumes ~1.48x, ~24x, ~2.5x, and ~34.7x lower energy per classification compared to conventional RF, SVM_RBF , MLP, and CNN, respectively. FoG is ~6.5x less energy efficient than SVM_LR, but achieves 18% higher accuracy on average across all considered datasets.
△ Less
Submitted 10 April, 2017;
originally announced April 2017.
-
Search for Screened Interactions Associated with Dark Energy Below the 100 $\mathrm{μm}$ Length Scale
Authors:
Alexander D. Rider,
David C. Moore,
Charles P. Blakemore,
Maxime Louis,
Marie Lu,
Giorgio Gratta
Abstract:
We present the results of a search for unknown interactions that couple to mass between an optically levitated microsphere and a gold-coated silicon cantilever. The scale and geometry of the apparatus enables a search for new forces that appear at distances below 100 $μ$m and which would have evaded previous searches due to screening mechanisms. The data are consistent with electrostatic backgroun…
▽ More
We present the results of a search for unknown interactions that couple to mass between an optically levitated microsphere and a gold-coated silicon cantilever. The scale and geometry of the apparatus enables a search for new forces that appear at distances below 100 $μ$m and which would have evaded previous searches due to screening mechanisms. The data are consistent with electrostatic backgrounds and place upper limits on the strength of new interactions at $<0.1$ fN in the geometry tested. For the specific example of a chameleon interaction with an inverse power law potential, these results exclude matter couplings $β> 5.6 \times 10^4$ in the region of parameter space where the self-coupling $Λ\gtrsim 5$ meV and the microspheres are not fully screened.
△ Less
Submitted 9 August, 2016; v1 submitted 17 April, 2016;
originally announced April 2016.